| *** noslzzp has joined #openstack-dev | 00:00 | |
| *** anniec has quit IRC | 00:00 | |
| *** tanisdl has quit IRC | 00:00 | |
| *** otherwiseguy has joined #openstack-dev | 00:01 | |
| *** glenng has quit IRC | 00:01 | |
| *** jobewan has quit IRC | 00:02 | |
| *** ayoung has joined #openstack-dev | 00:02 | |
| *** glenng has joined #openstack-dev | 00:02 | |
| *** epim has quit IRC | 00:03 | |
| *** reed has quit IRC | 00:04 | |
| *** ijw_ has quit IRC | 00:04 | |
| *** atiwari has quit IRC | 00:04 | |
| *** rcleere has quit IRC | 00:05 | |
| *** matsuhashi has joined #openstack-dev | 00:06 | |
| *** ijw_ has joined #openstack-dev | 00:08 | |
| *** anniec has joined #openstack-dev | 00:08 | |
| *** jobewan has joined #openstack-dev | 00:10 | |
| *** jobewan has quit IRC | 00:12 | |
| *** ijw_ has quit IRC | 00:13 | |
| *** wchrisj|away is now known as chris_johnson | 00:13 | |
| *** kmartin has quit IRC | 00:15 | |
| *** marcoemorais has quit IRC | 00:15 | |
| *** thomasem has quit IRC | 00:15 | |
| *** kmartin has joined #openstack-dev | 00:16 | |
| *** godara has joined #openstack-dev | 00:16 | |
| *** changbl has quit IRC | 00:16 | |
| *** marcoemorais has joined #openstack-dev | 00:16 | |
| *** ssurana has joined #openstack-dev | 00:16 | |
| *** epim has joined #openstack-dev | 00:16 | |
| *** ijw has joined #openstack-dev | 00:17 | |
| *** sweston has joined #openstack-dev | 00:19 | |
| *** ssurana has quit IRC | 00:19 | |
| *** ssurana has joined #openstack-dev | 00:20 | |
| *** matsuhashi has quit IRC | 00:21 | |
| *** marcoemorais has quit IRC | 00:21 | |
| *** morazi has quit IRC | 00:21 | |
| *** flaper87 is now known as flaper87|afk | 00:21 | |
| *** matsuhashi has joined #openstack-dev | 00:21 | |
| *** noslzzp has quit IRC | 00:21 | |
| *** comay has quit IRC | 00:23 | |
| *** buzztrol_ has joined #openstack-dev | 00:24 | |
| *** clayb has quit IRC | 00:25 | |
| *** matsuhashi has quit IRC | 00:25 | |
| *** matsuhas_ has joined #openstack-dev | 00:25 | |
| *** comay has joined #openstack-dev | 00:25 | |
| *** buzztroll has quit IRC | 00:25 | |
| *** noslzzp has joined #openstack-dev | 00:27 | |
| *** praneshp has quit IRC | 00:27 | |
| *** mkollaro has quit IRC | 00:28 | |
| *** pablosan has joined #openstack-dev | 00:28 | |
| *** mriedem has joined #openstack-dev | 00:28 | |
| *** kevinconway has quit IRC | 00:29 | |
| *** relaxdiego has joined #openstack-dev | 00:30 | |
| *** anniec has quit IRC | 00:30 | |
| *** radez is now known as radez_g0n3 | 00:30 | |
| *** anniec has joined #openstack-dev | 00:34 | |
| *** relaxdiego has quit IRC | 00:34 | |
| *** noslzzp has quit IRC | 00:34 | |
| *** thuc has joined #openstack-dev | 00:36 | |
| *** thuc has quit IRC | 00:37 | |
| *** noslzzp has joined #openstack-dev | 00:37 | |
| *** newell_ has joined #openstack-dev | 00:37 | |
| *** thuc has joined #openstack-dev | 00:38 | |
| *** pixelb has quit IRC | 00:39 | |
| *** abhirc has quit IRC | 00:39 | |
| *** ssurana has quit IRC | 00:41 | |
| *** noslzzp has quit IRC | 00:42 | |
| morganfainberg | bknudson, ooooh look at that | 00:42 |
|---|---|---|
| morganfainberg | bknudson, kds options are being picked up by the auto-gen tool | 00:42 |
| roaet | anyone familiar with tox asking for a password when attempting to build the environment? | 00:43 |
| *** ytwu has quit IRC | 00:45 | |
| morganfainberg | jamielennox, ping | 00:45 |
| jamielennox | morganfainberg: hello | 00:45 |
| morganfainberg | jamielennox, so i am about >.< close to proposing a fix so we can use auto-sample config generation | 00:45 |
| morganfainberg | jamielennox, the only issue left is that kds options are being picked up | 00:46 |
| jamielennox | morganfainberg: nice i tried that a while ago | 00:46 |
| jamielennox | hmmm, yea | 00:46 |
| morganfainberg | do you mind if i do a minor mangling of these options to work more like keystone's temporarily | 00:46 |
| jamielennox | oslo config is not designed to have multiple projects like this | 00:46 |
| morganfainberg | just make it a dict instead of a list | 00:46 |
| jamielennox | what specifically? | 00:46 |
| morganfainberg | instead of API_SERVICE_OPTS = [] | 00:47 |
| morganfainberg | jamielennox, API_SERVICE_OPTS = { None: [<opts>] } and have the registration happen based on that | 00:47 |
| roaet | nevermind, figured it out | 00:47 |
| morganfainberg | the config generator is silly | 00:47 |
| *** newell_ has quit IRC | 00:47 | |
| morganfainberg | jamielennox, it can't parse a dict | 00:47 |
| morganfainberg | jamielennox, it wont change the opts or how they get registered | 00:48 |
| morganfainberg | jamielennox, just changes if the generator can see them | 00:48 |
| *** newell has quit IRC | 00:48 | |
| morganfainberg | auto-config-generator | 00:48 |
| *** ytwu has joined #openstack-dev | 00:48 | |
| jamielennox | .., are you trying to hide KDS or do KDS as well? | 00:48 |
| *** newell has joined #openstack-dev | 00:48 | |
| morganfainberg | hide KDS | 00:48 |
| jamielennox | right, hide | 00:48 |
| morganfainberg | this is for keystone | 00:48 |
| morganfainberg | we'll split KDS off into it's own project soon i hope (erm git repo) | 00:49 |
| *** zhikunliu has joined #openstack-dev | 00:49 | |
| morganfainberg | then this doesn't matter | 00:49 |
| jamielennox | if that's the only way then ok, i assume you can't just exclude the directory | 00:49 |
| morganfainberg | nope | 00:50 |
| jamielennox | fix the generator to allow excluding a dir ? | 00:50 |
| morganfainberg | jamielennox, oslo-incubator... | 00:52 |
| morganfainberg | actually... let me just put help strings saying these are KDS and to ignore them. | 00:52 |
| morganfainberg | or | 00:52 |
| morganfainberg | argh | 00:52 |
| morganfainberg | ... | 00:52 |
| morganfainberg | asfksd | 00:52 |
| morganfainberg | fjadslkfj | 00:52 |
| *** sn6i23a has quit IRC | 00:53 | |
| jamielennox | seizure? | 00:53 |
| *** ramishra has quit IRC | 00:53 | |
| morganfainberg | lol | 00:53 |
| *** MaxV has joined #openstack-dev | 00:53 | |
| jamielennox | umm, if you are doing this as a tool can you do a script that just does rm -rf contrib/kds; gen_config; git checkout contrib/kds | 00:53 |
| *** epim has quit IRC | 00:53 | |
| morganfainberg | jamielennox, it looks for all .py files and extracts the OPTs | 00:54 |
| *** bdpayne has joined #openstack-dev | 00:54 | |
| morganfainberg | jamielennox, i'm just adding (KDS) in the front of the help text for your two options | 00:54 |
| *** praneshp has joined #openstack-dev | 00:54 | |
| morganfainberg | jamielennox, just indicates it's KDS specific | 00:54 |
| jamielennox | morganfainberg: ok, just trying to exhaust other options | 00:55 |
| morganfainberg | jamielennox, would need to break it out of the repo | 00:55 |
| *** tzabal has left #openstack-dev | 00:55 | |
| morganfainberg | jamielennox, it's really the only option | 00:55 |
| jamielennox | that would be good | 00:55 |
| *** bdpayne_ has quit IRC | 00:56 | |
| *** buzztroll has joined #openstack-dev | 00:57 | |
| *** buzztrol_ has quit IRC | 00:57 | |
| *** MaxV has quit IRC | 00:58 | |
| *** melwitt has quit IRC | 00:58 | |
| *** kgriffs is now known as kgriffs_afk | 00:59 | |
| *** pablosan has quit IRC | 00:59 | |
| *** xarses has quit IRC | 00:59 | |
| *** CaptTofu has joined #openstack-dev | 01:00 | |
| *** browne has quit IRC | 01:01 | |
| *** nosnos has joined #openstack-dev | 01:01 | |
| *** cybojanek has joined #openstack-dev | 01:02 | |
| *** dperaza has quit IRC | 01:03 | |
| *** carl_baldwin has quit IRC | 01:03 | |
| morganfainberg | jamielennox, yeah lets see about getting KDS into it's own repo asap :) | 01:03 |
| jamielennox | morganfainberg: i don't even know who to go to about that one | 01:03 |
| morganfainberg | jamielennox, i think we just need to go poke infra for a repo | 01:04 |
| morganfainberg | and make sure it lives under the keystone-core etc stuff | 01:04 |
| morganfainberg | oooor | 01:04 |
| morganfainberg | even have a KDS team group made *shrug* that is managed under 'identity' program | 01:04 |
| morganfainberg | yanno | 01:04 |
| morganfainberg | should poke dolphm about that again :) | 01:04 |
| *** sarob has joined #openstack-dev | 01:05 | |
| *** david-lyle has quit IRC | 01:05 | |
| *** IanGovett has quit IRC | 01:06 | |
| *** devlaps has quit IRC | 01:07 | |
| *** stevemar has joined #openstack-dev | 01:09 | |
| *** devoid has quit IRC | 01:10 | |
| *** xmltok has quit IRC | 01:11 | |
| *** noslzzp has joined #openstack-dev | 01:11 | |
| *** jdennis has quit IRC | 01:11 | |
| *** sarob has quit IRC | 01:12 | |
| *** nelsnelson has quit IRC | 01:12 | |
| morganfainberg | jamielennox, lol 13 files changed, 1995 insertions(+), 501 deletions(-), amazing what happens when you put a bunch more info in the sample config | 01:14 |
| jamielennox | wow, that's huge | 01:14 |
| morganfainberg | jamielennox, well the sample configuration now has _every_ option in it | 01:15 |
| morganfainberg | yes, that includes kombu etc | 01:15 |
| morganfainberg | jamielennox, and i had to update a ton of the options to have help texts | 01:15 |
| *** noslzzp has quit IRC | 01:15 | |
| jamielennox | morganfainberg: the problem i found with that is that there is a lot of options coming out of openstack.common that we don't actually use | 01:15 |
| morganfainberg | jamielennox, yep, and that isn't something i can really change | 01:15 |
| morganfainberg | jamielennox, so if we aren't using it, we should eliminate it | 01:16 |
| *** cybojanek has left #openstack-dev | 01:16 | |
| *** godara has quit IRC | 01:16 | |
| *** cybojanek has joined #openstack-dev | 01:17 | |
| *** anniec has quit IRC | 01:17 | |
| jamielennox | things like notifier though would pull in the whole dir | 01:17 |
| *** ykhodork has quit IRC | 01:17 | |
| morganfainberg | jamielennox, yep | 01:18 |
| jamielennox | we would have to start listing common by file | 01:18 |
| morganfainberg | jamielennox, but it properly reflects the options available in keystone | 01:18 |
| *** ctracey|away is now known as ctracey | 01:18 | |
| jamielennox | yea, it would be a good idea | 01:19 |
| *** dims has quit IRC | 01:21 | |
| *** unicell has joined #openstack-dev | 01:23 | |
| *** ramishra has joined #openstack-dev | 01:24 | |
| morganfainberg | jamielennox, https://review.openstack.org/#/c/72808/ | 01:24 |
| morganfainberg | bknudson, https://review.openstack.org/#/c/72808/ | 01:24 |
| *** RajeshMohan has quit IRC | 01:25 | |
| *** Edward-Zhang has joined #openstack-dev | 01:25 | |
| *** RajeshMohan has joined #openstack-dev | 01:26 | |
| *** abhirc_ has joined #openstack-dev | 01:27 | |
| *** jdennis has joined #openstack-dev | 01:27 | |
| *** krtaylor has joined #openstack-dev | 01:28 | |
| *** ramishra has quit IRC | 01:28 | |
| morganfainberg | bknudson, now with updated keystone.conf.sample https://review.openstack.org/#/c/72266/2 | 01:30 |
| morganfainberg | dhellmann_, thanks for the help today!! https://review.openstack.org/#/c/72808/ | 01:30 |
| *** zaitcev has quit IRC | 01:31 | |
| morganfainberg | dolphm, stable/havana fix ready for looking at https://review.openstack.org/#/c/66149/ | 01:32 |
| *** cagrev has joined #openstack-dev | 01:32 | |
| ayoung | dolphm, https://review.openstack.org/#/c/71181/ is very different now. Vast majority of that patch is sample data for tests. | 01:33 |
| *** Edward-Zhang has quit IRC | 01:34 | |
| *** mikeoutland has quit IRC | 01:35 | |
| nkinder | ayoung: I have a patch just about ready for review for eliminating extra searches for id_to_dn. | 01:35 |
| ayoung | nkinder, nice! | 01:35 |
| nkinder | ayoung: I was able to extend it groups and it's passing all of the tests. I cleaned up some of the user code too. | 01:35 |
| ayoung | no surprise there | 01:36 |
| ayoung | henrynash is going to need that code for the multi LDAP work | 01:36 |
| nkinder | ayoung: I'm guessing it's OK to punt on the other objects for assignment since we usually have that in sql? | 01:36 |
| ayoung | nkinder, it would be good to have both of them done, but I think two separate patches is appropriate | 01:37 |
| ayoung | nkinder, I think that was one of the comments in the last code review | 01:38 |
| *** thuc has quit IRC | 01:38 | |
| *** anniec has joined #openstack-dev | 01:38 | |
| nkinder | ayoung: ok, I'll double check that everything is working well. I did hook it up to 389 DS to verify that there is only one search when running some of the client commands | 01:38 |
| *** thuc has joined #openstack-dev | 01:38 | |
| ayoung | nkinder, so unit testing for LDAP is a little wonky, but if you can run the LDAP live tests it would be appreciated | 01:39 |
| nkinder | ayoung: somethign with the script to populate sample data wasn't working right when using the LDAP driver though. | 01:39 |
| ayoung | they might have bit rotted | 01:39 |
| ayoung | it used to be | 01:39 |
| ayoung | ./run_test.sh _ldap_livetests | 01:39 |
| nkinder | ayoung: well, devstack was broken for LDAP (which I fixed) | 01:39 |
| ayoung | yeah... | 01:39 |
| ayoung | nkinder, the goal when henrynash gets the multi domain backnends is to have the gate run with both SQL and LDAP for identity | 01:40 |
| nkinder | so _ldap_livetests just assumes what about the ldap config? Will it just use my keystone.conf settings? | 01:40 |
| ayoung | nkinder, it has its own config file | 01:40 |
| nkinder | ayoung: so it will set up openldap? | 01:40 |
| morganfainberg | nkinder, i think it uses the live test config in keystone/tests/ | 01:40 |
| ayoung | no | 01:41 |
| ayoung | it assumes an ldap server already running | 01:41 |
| ayoung | see backend_liveldap.conf | 01:41 |
| ayoung | its based on the devstack LDAP setup | 01:41 |
| nkinder | ok, which I need to configure in a certain way to match the tests I suppose | 01:41 |
| ayoung | but feel free to munge that file to run the tests | 01:41 |
| nkinder | ayoung: I'll likely just mimic the devstack setup with 389 DS then | 01:42 |
| ayoung | ++ | 01:42 |
| nkinder | ok, well that's work for tomorrow then (not like I had any already) ;) | 01:42 |
| jamielennox | bknudson: you've been doing the py33 work on client - does this look ok to you? https://review.openstack.org/#/c/72661/1/keystoneclient/middleware/auth_token.py | 01:43 |
| *** thuc has quit IRC | 01:43 | |
| ayoung | nkinder, BTW...I think I have a compressed token solution that should work around the Apache Headersize issues | 01:44 |
| ayoung | https://review.openstack.org/#/c/71181/ | 01:44 |
| *** tqtran has quit IRC | 01:44 | |
| ayoung | and with that...I'm going to go get dinner | 01:44 |
| nkinder | ayoung: enjoy | 01:45 |
| *** _cjones_ has quit IRC | 01:45 | |
| ayoung | yeah | 01:45 |
| *** tris has quit IRC | 01:48 | |
| *** mgagne has quit IRC | 01:49 | |
| *** lbragstad has joined #openstack-dev | 01:49 | |
| *** ayoung has quit IRC | 01:50 | |
| *** chris_johnson is now known as wchrisj|away | 01:51 | |
| *** danielbruno has quit IRC | 01:51 | |
| morganfainberg | jamielennox, that is my understanding of the proper py33 way | 01:53 |
| *** tsekiyam_ has joined #openstack-dev | 01:53 | |
| *** jckasper has joined #openstack-dev | 01:53 | |
| jamielennox | morganfainberg: yea i thought so | 01:53 |
| *** muehsi has quit IRC | 01:53 | |
| jamielennox | morganfainberg: i hate string encoding | 01:54 |
| morganfainberg | jamielennox, ++++++++++++++++++ | 01:54 |
| morganfainberg | jamielennox, i loathe it | 01:54 |
| jamielennox | i don't care if py3 is better or worse i just never want to deal with it | 01:54 |
| morganfainberg | you need to deal with it some limited cases in py3 | 01:55 |
| *** killer_prince has quit IRC | 01:55 | |
| *** jckasper has quit IRC | 01:55 | |
| morganfainberg | when you are dealing with hashing (needs to be byte_str) | 01:55 |
| morganfainberg | but it's super limited | 01:55 |
| *** jckasper has joined #openstack-dev | 01:55 | |
| morganfainberg | i think i've run into an issue with it 2 or 3 times total. vs. every time someone uses a unicode text type... utf-8 | 01:55 |
| morganfainberg | oh yay! | 01:56 |
| jamielennox | morganfainberg: i was doing some socket level stuff and it was a right PITA | 01:56 |
| morganfainberg | review day now guarantees a WIP patchset has a score of 0 | 01:56 |
| morganfainberg | woooo :) | 01:57 |
| *** tsekiyama has quit IRC | 01:57 | |
| morganfainberg | so much better | 01:57 |
| morganfainberg | now it claimes youre review Add cryptographic key storage is the #1 review for keystone | 01:57 |
| *** tsekiyam_ has quit IRC | 01:58 | |
| morganfainberg | s/youre/your | 01:58 |
| morganfainberg | jamielennox, ^ | 01:58 |
| jamielennox | lol, yea, it's been up for ages | 01:58 |
| jamielennox | but there are dependants | 01:58 |
| morganfainberg | but it now is the top of the list | 01:58 |
| morganfainberg | jamielennox, because before ayoung's WIPs were all the top | 01:58 |
| morganfainberg | the ones.. from like pre-havana release | 01:59 |
| jamielennox | morganfainberg: hmm, it really should ignore WIPs | 01:59 |
| morganfainberg | jamielennox, it effectively does now. i don't think they shouldn't appear | 01:59 |
| *** devoid has joined #openstack-dev | 01:59 | |
| morganfainberg | jamielennox, maybe i should also propose a change that makes any non-wip score+1 value | 02:00 |
| morganfainberg | so WIPs are all at the bottom | 02:00 |
| morganfainberg | actually... i think i might do that. | 02:00 |
| *** alop has quit IRC | 02:02 | |
| *** xuhanp has joined #openstack-dev | 02:02 | |
| jamielennox | i havene't used reviewday, there was a next-review or something | 02:02 |
| *** anniec has quit IRC | 02:03 | |
| *** amcrn has quit IRC | 02:03 | |
| morganfainberg | jamielennox, as soon as the JSON becomes available i will be making next-review use review-day's json (optionally) | 02:05 |
| *** dkranz has joined #openstack-dev | 02:06 | |
| morganfainberg | jamielennox, http://status.openstack.org/reviews/reviewday.json :( | 02:07 |
| *** sweston has quit IRC | 02:07 | |
| jamielennox | Forbidden | 02:07 |
| morganfainberg | yeah | 02:07 |
| morganfainberg | as soon as that is fixed | 02:07 |
| morganfainberg | next review can be even coooooooler | 02:07 |
| *** sarob has joined #openstack-dev | 02:09 | |
| *** tkay has quit IRC | 02:12 | |
| *** sarob has quit IRC | 02:13 | |
| *** yaguang has joined #openstack-dev | 02:14 | |
| *** gongysh has joined #openstack-dev | 02:18 | |
| noorul | hello all | 02:18 |
| *** MaxV has joined #openstack-dev | 02:19 | |
| noorul | How do we write plugins for keystone | 02:19 |
| noorul | Is there any documentations? | 02:19 |
| noorul | s/is/are | 02:19 |
| *** sweston has joined #openstack-dev | 02:20 | |
| *** muehsi has joined #openstack-dev | 02:20 | |
| noorul | I am looking to have separate identity and assignment back-end | 02:21 |
| *** tdruiva has joined #openstack-dev | 02:22 | |
| *** spzala has quit IRC | 02:23 | |
| *** MaxV has quit IRC | 02:23 | |
| *** ramishra has joined #openstack-dev | 02:24 | |
| *** erkules_ has joined #openstack-dev | 02:26 | |
| *** Brainspackle has quit IRC | 02:26 | |
| *** Brainspackle has joined #openstack-dev | 02:26 | |
| *** zoresvit has joined #openstack-dev | 02:27 | |
| *** erkules has quit IRC | 02:28 | |
| *** ramishra_ has joined #openstack-dev | 02:28 | |
| *** ramishra has quit IRC | 02:29 | |
| jamielennox | noorul: essentially no, you'll have to have a look at the drivers in keystone and try to emulate them | 02:30 |
| *** angdraug has quit IRC | 02:33 | |
| *** pradeep has joined #openstack-dev | 02:35 | |
| *** novas0x2a|laptop has quit IRC | 02:38 | |
| *** godara has joined #openstack-dev | 02:39 | |
| *** ramishra_ has quit IRC | 02:40 | |
| noorul | jamielennox: Can I have this code in separate repository? | 02:41 |
| jamielennox | noorul: yes that's fine | 02:41 |
| noorul | jamielennox: Python should be able to import the package, right? | 02:41 |
| jamielennox | if you look at the backend config it's of the form driver = keystone.identity.backends.sql.Identity | 02:41 |
| *** galstrom_zzz is now known as galstrom | 02:41 | |
| jamielennox | if you match that structure then it will load it for you | 02:42 |
| noorul | So I should have keystone/identity/backends directory | 02:42 |
| *** godara has quit IRC | 02:43 | |
| *** godara has joined #openstack-dev | 02:43 | |
| *** buzztroll has quit IRC | 02:43 | |
| noorul | for example if i have xx.py then I can just replace config to looke like this | 02:44 |
| *** erkules_ has quit IRC | 02:44 | |
| *** buzztroll has joined #openstack-dev | 02:44 | |
| noorul | driver = keystone.identity.backends.xx.Identity | 02:44 |
| *** mwagner_lap has joined #openstack-dev | 02:45 | |
| *** anniec has joined #openstack-dev | 02:45 | |
| *** erkules_ has joined #openstack-dev | 02:46 | |
| *** Tross has joined #openstack-dev | 02:46 | |
| *** godara_ has joined #openstack-dev | 02:46 | |
| *** godara has quit IRC | 02:48 | |
| jamielennox | yes | 02:48 |
| jamielennox | noorul: or you can develop it as a package outside of keystone and install it and it will load that to | 02:49 |
| *** buzztroll has quit IRC | 02:49 | |
| *** stevemar has quit IRC | 02:51 | |
| jamielennox | all hail Steap_ keystoneclient's first pass of the py33 gate: https://review.openstack.org/#/c/72661/ | 02:51 |
| *** dperaza has joined #openstack-dev | 02:51 | |
| jamielennox | bknudson, morganfainberg, gyee, dolphm: ^^ | 02:52 |
| morganfainberg | jamielennox, wooooooooo | 02:52 |
| morganfainberg | jamielennox, guess that means we need to make the py33 gate for keystoneclient voting | 02:53 |
| morganfainberg | jamielennox, (once that merges) | 02:53 |
| jamielennox | morganfainberg, asap | 02:53 |
| jamielennox | i'm pretty sure he's on it | 02:54 |
| *** Mandell has quit IRC | 02:54 | |
| *** cagrev has quit IRC | 02:54 | |
| *** achampion has quit IRC | 02:56 | |
| *** xarses has joined #openstack-dev | 02:56 | |
| morganfainberg | jamielennox, looks like changing it here: https://github.com/openstack-infra/config/blob/master/modules/openstack_project/files/zuul/layout.yaml#L441 and adding python-keystonecleint-33 to the gate block https://github.com/openstack-infra/config/blob/master/modules/openstack_project/files/zuul/layout.yaml#L1308 should do it | 02:56 |
| *** ijw_ has joined #openstack-dev | 02:57 | |
| jamielennox | morganfainberg: he must be offline, i asked if he said he knows how and so i think we give him the honours | 02:57 |
| morganfainberg | jamielennox, oh totaly fine with him doing so | 02:58 |
| morganfainberg | jamielennox, just was looking to see what would do it :) | 02:58 |
| jamielennox | morganfainberg: anyway very excited for that change | 02:58 |
| morganfainberg | :) | 02:58 |
| jamielennox | sick of having to remember to do six | 02:58 |
| morganfainberg | lol | 02:58 |
| morganfainberg | you'll still need to remember | 02:58 |
| *** anniec has quit IRC | 02:58 | |
| *** ijw has quit IRC | 02:59 | |
| *** arnaud__ has quit IRC | 03:00 | |
| *** henrynash has joined #openstack-dev | 03:00 | |
| *** anniec has joined #openstack-dev | 03:03 | |
| *** tsekiyama has joined #openstack-dev | 03:04 | |
| *** arnaud__ has joined #openstack-dev | 03:05 | |
| *** godara_ has quit IRC | 03:07 | |
| *** ramishra has joined #openstack-dev | 03:08 | |
| *** tsekiyama has quit IRC | 03:08 | |
| *** sarob has joined #openstack-dev | 03:09 | |
| *** ykhodork has joined #openstack-dev | 03:09 | |
| *** marcoemorais has joined #openstack-dev | 03:12 | |
| *** ramishra has quit IRC | 03:13 | |
| *** gokrokve has joined #openstack-dev | 03:14 | |
| *** marcoemorais1 has joined #openstack-dev | 03:14 | |
| *** sarob has quit IRC | 03:16 | |
| *** marcoemorais has quit IRC | 03:16 | |
| *** annegentleweb has joined #openstack-dev | 03:17 | |
| *** rcleere has joined #openstack-dev | 03:17 | |
| *** sarob has joined #openstack-dev | 03:19 | |
| *** MaxV has joined #openstack-dev | 03:20 | |
| *** galstrom is now known as galstrom_zzz | 03:21 | |
| *** paragan has joined #openstack-dev | 03:21 | |
| *** paragan has quit IRC | 03:21 | |
| *** paragan has joined #openstack-dev | 03:21 | |
| *** achampion has joined #openstack-dev | 03:22 | |
| *** rraja has joined #openstack-dev | 03:23 | |
| *** praneshp has quit IRC | 03:23 | |
| *** sarob has quit IRC | 03:24 | |
| *** MaxV has quit IRC | 03:24 | |
| *** mlegault has joined #openstack-dev | 03:26 | |
| *** wchrisj|away is now known as chris_johnson | 03:26 | |
| *** annegentleweb has quit IRC | 03:26 | |
| *** chris_johnson has quit IRC | 03:27 | |
| *** mlegault_ has joined #openstack-dev | 03:27 | |
| *** CaptTofu has quit IRC | 03:31 | |
| *** mlegault has quit IRC | 03:31 | |
| *** jmontemayor has joined #openstack-dev | 03:31 | |
| *** arnaud__ has quit IRC | 03:31 | |
| *** arnaud has quit IRC | 03:31 | |
| *** galstrom_zzz is now known as galstrom | 03:32 | |
| *** byeager has joined #openstack-dev | 03:34 | |
| *** mlegault_ has quit IRC | 03:35 | |
| *** hdd_ has joined #openstack-dev | 03:36 | |
| *** matsuhas_ has quit IRC | 03:39 | |
| *** amerine_ has joined #openstack-dev | 03:40 | |
| *** amerine has quit IRC | 03:40 | |
| *** aditirav has joined #openstack-dev | 03:43 | |
| *** mgagne has joined #openstack-dev | 03:43 | |
| *** enykeev has joined #openstack-dev | 03:45 | |
| *** mikeoutland has joined #openstack-dev | 03:50 | |
| *** chandan_kumar has joined #openstack-dev | 03:51 | |
| *** baoli has joined #openstack-dev | 03:53 | |
| *** mgagne1 has joined #openstack-dev | 03:53 | |
| *** doug_shelley66 has quit IRC | 03:54 | |
| *** ykhodork has quit IRC | 03:54 | |
| *** buzztroll has joined #openstack-dev | 03:54 | |
| *** mgagne has quit IRC | 03:56 | |
| *** markmcclain has joined #openstack-dev | 03:56 | |
| *** praneshp has joined #openstack-dev | 03:56 | |
| *** ssurana has joined #openstack-dev | 03:56 | |
| *** markmcclain has quit IRC | 03:57 | |
| *** yamahata has quit IRC | 03:59 | |
| *** buzztroll has quit IRC | 03:59 | |
| *** dperaza has quit IRC | 03:59 | |
| *** dkehn__ has joined #openstack-dev | 04:00 | |
| *** rraja has quit IRC | 04:01 | |
| *** mgagne has joined #openstack-dev | 04:01 | |
| *** baoli has quit IRC | 04:01 | |
| *** mgagne1 has quit IRC | 04:03 | |
| *** dkehn_ has quit IRC | 04:03 | |
| *** arborism has joined #openstack-dev | 04:04 | |
| *** changbl has joined #openstack-dev | 04:04 | |
| *** yamahata has joined #openstack-dev | 04:05 | |
| *** amcrn_ has joined #openstack-dev | 04:05 | |
| Steap_ | jamielennox: thanks for letting me commit the last patch :) | 04:07 |
| *** byeager has quit IRC | 04:07 | |
| *** galstrom is now known as galstrom_zzz | 04:07 | |
| Steap_ | https://review.openstack.org/#/c/72831/ | 04:07 |
| jamielennox | Steap_: was nothing, just preventing morganfainberg getting to excited | 04:08 |
| Steap_ | hehe | 04:08 |
| *** byeager has joined #openstack-dev | 04:08 | |
| *** amcrn_ is now known as amcrn | 04:08 | |
| *** tanisdl has joined #openstack-dev | 04:08 | |
| jamielennox | Steap_: i have no idea of how infra works though so i won't vote on it | 04:08 |
| morganfainberg | Steap_. you want to revise that patch, make the "voting" line say true, and then add it like you did on 1308 | 04:08 |
| Steap_ | jamielennox: morganfainberg of the line is removed, isn't it true by default ? | 04:09 |
| morganfainberg | Steap_, afaict you removed the "template" for the python-keystoneclient-py33 gate | 04:09 |
| morganfainberg | Steap_, hm. maybe | 04:09 |
| *** arborism has quit IRC | 04:09 | |
| morganfainberg | Steap_, let me look at that. | 04:09 |
| Steap_ | git grep "voting: true" | 04:09 |
| Steap_ | returns nothing | 04:09 |
| morganfainberg | hehe | 04:09 |
| morganfainberg | ok then | 04:09 |
| morganfainberg | :) | 04:09 |
| morganfainberg | you did more research than i did! | 04:09 |
| morganfainberg | Steap_, but in either case, bravo on the py33 work!! | 04:09 |
| Steap_ | morganfainberg: no problem | 04:09 |
| Steap_ | morganfainberg: I was asking jamielennox in private earlier, but we might as well talk about this here | 04:10 |
| Steap_ | Many clients depend on keystoneclient, so I currently can't make them work with Python3 | 04:10 |
| morganfainberg | Steap_, oh no, what have i done, i proved i was alive and paying attention >.> | 04:10 |
| morganfainberg | Steap_, sure. | 04:11 |
| Steap_ | morganfainberg: do you think keystoneclient could have one more release a bit before IceHouse ? | 04:11 |
| jamielennox | WTF glanceclient - still using httplib! | 04:11 |
| jamielennox | sorry, carry on | 04:11 |
| Steap_ | that'd leave me enough time to port some of the other clients | 04:11 |
| morganfainberg | Steap_, ideally, yes. and likely we absolutely want it if revocation events get added into auth_token | 04:11 |
| Steap_ | (for instance, the only blocker for the Ceilometerclient is keystoneclient) | 04:11 |
| Steap_ | morganfainberg: ok | 04:12 |
| morganfainberg | Steap_, but i wouldn't see an issue (though dolphm may say otherwise) on getting a keystoneclient out now and then one closer to Icehouse landing | 04:12 |
| Steap_ | ok | 04:12 |
| Steap_ | Should I open a bug report so that everyone has time to discuss this ? | 04:12 |
| *** dkehn__ is now known as dkehn_ | 04:12 | |
| jamielennox | morganfainberg: i have some patches around auth_plugins i'm not sure would be breaking | 04:12 |
| morganfainberg | jamielennox, merged? | 04:13 |
| jamielennox | morganfainberg: no but not far i think | 04:13 |
| morganfainberg | jamielennox, or pending? or... ? | 04:13 |
| morganfainberg | jamielennox, ah, maybe get this released before that work? | 04:13 |
| jamielennox | i was thinking i had some more time with this release because we just did one | 04:13 |
| jamielennox | morganfainberg: yea not sure | 04:13 |
| jamielennox | morganfainberg: just have to see what i can support in future | 04:13 |
| morganfainberg | jamielennox, likely we can call this a minor point release since it's really for the py33 stuff | 04:14 |
| jamielennox | morganfainberg: i don't think that makes a difference for api | 04:14 |
| morganfainberg | jamielennox, fair enough | 04:14 |
| *** galstrom_zzz is now known as galstrom | 04:15 | |
| morganfainberg | jamielennox, i'd advocate for an extra release wedged in here if this really is the only blocker for some other py33 work getting into icehouse | 04:15 |
| jamielennox | morganfainberg: yea i agree py33 is worth messing around with the tree a bit | 04:15 |
| Steap_ | morganfainberg: well, maybe not the only one, but it's a blocker for mostclients :) | 04:15 |
| jamielennox | even if i have to revert a patch or two then put it back after | 04:15 |
| morganfainberg | Steap_, sure sure. but yes not blocking is good | 04:16 |
| morganfainberg | jamielennox, might be worth it. | 04:16 |
| morganfainberg | jamielennox, prob should poke at dolphm tomorrow and discuss it | 04:16 |
| *** ssurana has quit IRC | 04:16 | |
| jamielennox | yep | 04:16 |
| morganfainberg | similarly see about KDS. | 04:16 |
| morganfainberg | cause i really think we need to split that off | 04:16 |
| morganfainberg | jamielennox, will massively ease up a number of things that way | 04:17 |
| morganfainberg | testing, etc | 04:17 |
| *** godara has joined #openstack-dev | 04:18 | |
| *** ssurana has joined #openstack-dev | 04:19 | |
| *** SumitNaiksatam has joined #openstack-dev | 04:20 | |
| *** MaxV has joined #openstack-dev | 04:21 | |
| *** alop has joined #openstack-dev | 04:22 | |
| *** MaxV has quit IRC | 04:25 | |
| *** ssurana has quit IRC | 04:26 | |
| *** ssurana has joined #openstack-dev | 04:26 | |
| *** anniec has quit IRC | 04:27 | |
| *** ssurana has quit IRC | 04:28 | |
| *** byeager has quit IRC | 04:28 | |
| *** ssurana has joined #openstack-dev | 04:28 | |
| *** morazi has joined #openstack-dev | 04:31 | |
| *** jay-lau-513 has joined #openstack-dev | 04:33 | |
| *** killer_prince has joined #openstack-dev | 04:33 | |
| *** aeperezt has quit IRC | 04:34 | |
| *** mikeoutland has quit IRC | 04:34 | |
| *** mikeoutland has joined #openstack-dev | 04:35 | |
| *** schwicht has quit IRC | 04:35 | |
| *** Mandell has joined #openstack-dev | 04:36 | |
| *** jmontemayor has quit IRC | 04:37 | |
| *** lcheng_ has joined #openstack-dev | 04:38 | |
| *** ssurana has left #openstack-dev | 04:40 | |
| lbragstad | jamielennox: got rid of ['media-types'] right? | 04:41 |
| lbragstad | just looking through your new patch | 04:41 |
| jamielennox | lbragstad: just then | 04:41 |
| lbragstad | ok | 04:41 |
| morganfainberg | lbragstad, DOH! i can't believe i missed Memcached vs memcached (again) | 04:41 |
| lbragstad | :) | 04:41 |
| morganfainberg | lbragstad, thanks for the comment, brains a little fried | 04:41 |
| morganfainberg | lbragstad, https://review.openstack.org/#/c/72808/ - that fried my brain quite a bit :( | 04:41 |
| lbragstad | morganfainberg: it happnes... took me five patch sets to get my commit message right last night | 04:41 |
| morganfainberg | heh | 04:42 |
| morganfainberg | ouch | 04:42 |
| morganfainberg | :( | 04:42 |
| * morganfainberg doesn't envy that | 04:42 | |
| lbragstad | morganfainberg: you got that to work! | 04:42 |
| *** AlanClark has quit IRC | 04:42 | |
| morganfainberg | so i think i have a change, jamielennox that would add the KDS repo and hook in all the relevant stuff | 04:42 |
| morganfainberg | jamielennox, i'll ping dolph tomorrow before submitting it | 04:42 |
| lbragstad | I ended up getting a little swamped this week but I wanted to get back to the, I had a change up for that too. | 04:43 |
| morganfainberg | lbragstad, yesh, yesh i did. was a pain to add all the help lines for it to our config | 04:43 |
| jamielennox | morganfainberg: oh? ok - sounds good | 04:43 |
| lbragstad | morganfainberg: for some reason I couldn't get it to generate the options in keystone/common/config.py but I didn't have much time | 04:43 |
| morganfainberg | lbragstad, the hardest part was the stuff dhellmann_ was awesome about and so i got it to work similar to how oslo.messaging does by using pbr and all that | 04:43 |
| *** harlowja is now known as harlowja_away | 04:44 | |
| morganfainberg | lbragstad, yeah look at the setup.cfg and the change to keystone.common.config (bottom) to see what i did | 04:44 |
| morganfainberg | oh and the oslo.config.generator.rc in tools/config | 04:44 |
| lbragstad | oh... yeah... that makes sense | 04:45 |
| lbragstad | this is awesome, Brant and I have been talking about doing this | 04:45 |
| morganfainberg | could do it differently, but that was the path of least resistance | 04:45 |
| morganfainberg | and.. makes oslo sync much easier to chace options down. it also shows we might want to trim down the openstack.common modules :P | 04:45 |
| lbragstad | not that it probably matters, but giving you credit here and pointing to your change: https://review.openstack.org/#/c/67595/ | 04:46 |
| *** jkoelker has quit IRC | 04:47 | |
| *** jkoelker has joined #openstack-dev | 04:47 | |
| morganfainberg | lbragstad, oh hehe didn't even see it | 04:48 |
| lbragstad | it was WIP for a while | 04:48 |
| *** tanisdl has quit IRC | 04:48 | |
| morganfainberg | lbragstad aghh | 04:48 |
| morganfainberg | ahh* | 04:48 |
| morganfainberg | lbragstad, and i am working on solving the KDS issue ;) | 04:48 |
| morganfainberg | issue being that it should be in it's own repo :) | 04:49 |
| jamielennox | morganfainberg, lbragstad: i'm working on some discovery stuff and would like some opinions if you have a sec | 04:49 |
| lbragstad | so, qq on https://review.openstack.org/#/c/72808/3/etc/keystone.conf.sample | 04:49 |
| lbragstad | jamielennox: sure, same patch as the media-types one? | 04:49 |
| jamielennox | lbragstad: completely unrelated | 04:50 |
| lbragstad | ahh | 04:50 |
| lbragstad | morganfainberg: line 38 from the old sample config, there is a note from dolphm isn't that something that should move to oslo? | 04:50 |
| morganfainberg | jamielennox, i knew i should have stayed quiet. | 04:50 |
| *** nacim has joined #openstack-dev | 04:50 | |
| jamielennox | given what we know about keystone auth, and ignoring the client that we have what does a good auth system look like | 04:50 |
| jamielennox | i've been thinking on it for a while but i'm stuck on bridging the v2/v3 apis in an obvious way | 04:51 |
| morganfainberg | lbragstad, uhm. | 04:51 |
| * lbragstad digs... | 04:51 | |
| lbragstad | I had this thought out before | 04:51 |
| lbragstad | I can't remember where I put it | 04:51 |
| morganfainberg | lbragstad, could add that in the help text. | 04:51 |
| lbragstad | ++ | 04:52 |
| lbragstad | just so we don't lose the note for now | 04:52 |
| morganfainberg | lbragstad, if you think it's important to capture, please comment (point it out in the config.py stuff) | 04:52 |
| *** matsuhashi has joined #openstack-dev | 04:52 | |
| morganfainberg | lbragstad, i am not going to push anothe rpatchset for that tonight, want to circle back w/ bknudson, ayoung, dolphm etc and see if we _Really_ want this now | 04:52 |
| lbragstad | yeah, good idea. | 04:53 |
| lbragstad | morganfainberg: thinking early J for that? | 04:53 |
| jamielennox | morganfainberg: early J might be a good idea | 04:53 |
| morganfainberg | lbragstad, i'd like it now, but honestly, it might just be a bit of a bear | 04:53 |
| *** rraja has joined #openstack-dev | 04:53 | |
| jamielennox | lets us trim down the config options before we do a release | 04:53 |
| morganfainberg | lbragstad, it would be really really good to not having to manage the config sample anymore before the release | 04:54 |
| morganfainberg | lbragstad, but it might go for J instead. | 04:54 |
| morganfainberg | jamielennox, so. back to auth | 04:54 |
| jamielennox | morganfainberg: :) | 04:54 |
| *** markwash has joined #openstack-dev | 04:55 | |
| jamielennox | So i'm pretty sure 1 plugin per auth method is the right way to do it | 04:55 |
| morganfainberg | jamielennox, i would say yes | 04:55 |
| *** henrynash has quit IRC | 04:55 | |
| jamielennox | and there is enough of an overlap between all v2 plugins and all v3 plugins that i can somewhat group them together | 04:55 |
| jamielennox | not that i tthink that's necessarily a good idea | 04:56 |
| *** armax has joined #openstack-dev | 04:56 | |
| *** armax has quit IRC | 04:56 | |
| morganfainberg | lbragstad, that option, i think comes from oslo-incubator. and therefore we can't add to the help string. ion keystone | 04:56 |
| jamielennox | but there are fundamental differences between v2 and v3 that i've no idea if it's possible to get over that | 04:56 |
| morganfainberg | jamielennox, no, i'd make V2 and V3 plugins separate... | 04:56 |
| morganfainberg | jamielennox, actually, i'd make V2 a single plugin. | 04:57 |
| morganfainberg | jamielennox, and V3 can be split up | 04:57 |
| morganfainberg | jamielennox, v2 is dead in K. | 04:57 |
| jamielennox | morganfainberg: well i did v2 anyway because it was trivial | 04:57 |
| lbragstad | morganfainberg: yep in openstack/common/policy.py | 04:57 |
| jamielennox | I'm just not sure what the CLI interaction is for example | 04:57 |
| morganfainberg | jamielennox, I'd say keep v2 completly separate, just let V2 live isolated from v3 if at all possible | 04:57 |
| lbragstad | that is where it's registered, so moving that to [policy] section will break policy.py | 04:57 |
| jamielennox | i don't think you should have to say v2 or v3 auth | 04:57 |
| jamielennox | morganfainberg: ++ | 04:58 |
| *** lcheng_ has quit IRC | 04:58 | |
| morganfainberg | jamielennox, i think you should default to the highest supported version and allow the cli to say --version-2 | 04:58 |
| morganfainberg | jamielennox, you are doing discovery, right? | 04:58 |
| morganfainberg | jamielennox, (that was rhetorical) | 04:58 |
| morganfainberg | jamielennox, so if the client can do V3 and V3 is available, do v3 unless told otherwise | 04:58 |
| morganfainberg | jamielennox, so if you want V2 auth, you would need to specify it, since it is not the highest "Stable" api | 04:59 |
| jamielennox | ok, so maybe another question, how do you ask for a version | 04:59 |
| morganfainberg | jamielennox, rephrase, default to highest stable. | 04:59 |
| morganfainberg | jamielennox, ask keystone, right? | 04:59 |
| jamielennox | hmm, not really | 04:59 |
| *** nacim has quit IRC | 04:59 | |
| morganfainberg | jamielennox, we say what the highest stable version the server supports is. | 04:59 |
| morganfainberg | jamielennox, on / | 04:59 |
| jamielennox | morganfainberg: yea, the problem is the bootstrapping | 05:00 |
| morganfainberg | jamielennox, could you leverage that? | 05:00 |
| jamielennox | once i've got a service catalog it's fine | 05:00 |
| morganfainberg | jamielennox, same rules apply | 05:00 |
| jamielennox | but what does a good interaction with the initial discover look like | 05:00 |
| morganfainberg | jamielennox, ask the server for versions, ask the same server for catalog/auth/whatever is needed | 05:00 |
| morganfainberg | jamielennox, AUTH_URL would point at like http://keystone:35357/ | 05:01 |
| morganfainberg | that would return "i talk V2, V3, V4, V4.5" | 05:01 |
| morganfainberg | or some such | 05:01 |
| morganfainberg | v4 is not stable, v3 is, v2 is stable | 05:01 |
| jamielennox | right i've kind of got that alreday | 05:01 |
| morganfainberg | v3 is highest stable, so ask the same server v3 auth | 05:01 |
| morganfainberg | and/or catalog etc | 05:01 |
| *** edmund has quit IRC | 05:01 | |
| * morganfainberg makes mental note that being able to use srv records in dns would be super cool | 05:02 | |
| morganfainberg | instead of AUTH_URL env var, et al | 05:02 |
| jamielennox | dns has up before | 05:02 |
| jamielennox | has come up before | 05:02 |
| morganfainberg | i'd like to support it in general | 05:02 |
| morganfainberg | but not make it the only way to do it | 05:02 |
| morganfainberg | but ehh. | 05:02 |
| morganfainberg | moving on | 05:02 |
| *** kbrierly has quit IRC | 05:03 | |
| morganfainberg | if AUTH_URL is http://keystone:37257/v2.0/.... | 05:03 |
| *** mikeoutland has quit IRC | 05:03 | |
| morganfainberg | i mean.. huh | 05:03 |
| *** galstrom is now known as galstrom_zzz | 05:03 | |
| morganfainberg | you need a way to say don't discover | 05:04 |
| *** zul has joined #openstack-dev | 05:04 | |
| morganfainberg | and you need a way to say "use something besides latest stable" | 05:04 |
| jamielennox | morganfainberg: so like https://etherpad.openstack.org/p/version-discover | 05:04 |
| morganfainberg | sure | 05:04 |
| jamielennox | it just seems a bit too low level | 05:04 |
| morganfainberg | maybe sec. | 05:05 |
| *** kushal has joined #openstack-dev | 05:05 | |
| *** morazi has quit IRC | 05:05 | |
| morganfainberg | maybe like that? | 05:06 |
| morganfainberg | or similar? | 05:06 |
| jamielennox | what does newest_stable return though/ a string? | 05:07 |
| morganfainberg | i'd make it a factory | 05:07 |
| jamielennox | ok, so the problem is somewhat that we don't want to make a client | 05:07 |
| jamielennox | the discovery that i've implemented kind of does that - it was a partial mistake | 05:08 |
| *** yuan has quit IRC | 05:08 | |
| *** godara has quit IRC | 05:08 | |
| jamielennox | the very first thing we need to do is create an auth mechanism | 05:08 |
| morganfainberg | sure | 05:08 |
| *** zul has quit IRC | 05:08 | |
| morganfainberg | it could be a stubby factory that has the V3 or V2 or whatever magic logic is needed to move beyond the auth method | 05:08 |
| jamielennox | so the factory concept is a bit more difficult in that case because auth mechanisms are a variety of plugins | 05:08 |
| morganfainberg | it's not really a client, it's the auth_object | 05:09 |
| morganfainberg | the factory could then load/manage the auth_plugin side | 05:09 |
| *** gokrokve has quit IRC | 05:09 | |
| morganfainberg | the way i see it is auth_object is then applied to the client. | 05:10 |
| morganfainberg | so this first step is getting the auth_object and all stuff along with it. | 05:10 |
| jamielennox | right that's pretty much what i'm going for | 05:10 |
| morganfainberg | it's then wrapped into the client when you create your client object | 05:10 |
| jamielennox | but there are two types of question then | 05:10 |
| *** ramishra has joined #openstack-dev | 05:10 | |
| morganfainberg | the stubby auth thing (in my semi-fried brain) should handle the auth_plugin stuff | 05:11 |
| *** galstrom_zzz is now known as galstrom | 05:11 | |
| jamielennox | three: what is the most recent version, is this version supported, give me the factory to create aut for this | 05:11 |
| morganfainberg | the client wont care, it knows what the auth entry point is. | 05:11 |
| morganfainberg | jamielennox, sure that seems correct. | 05:11 |
| jamielennox | my brain is fried too, concert last night followed by early morning meeting means not at my best | 05:11 |
| lbragstad | jamielennox: nice, good show? | 05:12 |
| jamielennox | the national, yea they were good | 05:12 |
| morganfainberg | jamielennox, if the discover object is always responsible for discovery, in all ways, you can ask the first two questions easily | 05:12 |
| jamielennox | right | 05:12 |
| jamielennox | I'm not sure to the extent we are supporting this factory method for auth | 05:13 |
| morganfainberg | jamielennox, perhaps discovery and auth are really one thing | 05:13 |
| *** cadenzajon has joined #openstack-dev | 05:13 | |
| *** sarob has joined #openstack-dev | 05:13 | |
| jamielennox | i've had it there mostly for the CLI and CONF cases | 05:13 |
| jamielennox | morganfainberg: i don't think so, because auth can take on so many forms | 05:14 |
| morganfainberg | jamielennox, ok. | 05:14 |
| jamielennox | besides i prefer the idea of making primatives and then combining them into higher level interfaces | 05:14 |
| morganfainberg | jamielennox, ++ | 05:14 |
| jamielennox | if you need to use the primatives they are there | 05:14 |
| *** galstrom is now known as galstrom_zzz | 05:15 | |
| morganfainberg | jamielennox, yeah, brain is ... not here | 05:15 |
| morganfainberg | i looked at auto generating config stuff today...and wrote help strings | 05:15 |
| morganfainberg | well.. copy/pasted/reformatted | 05:15 |
| morganfainberg | :P | 05:15 |
| jamielennox | that's not easy, i looked through that conf generation before | 05:16 |
| jamielennox | actually i filed a bug for it against keystone ages ago | 05:16 |
| jamielennox | you might close that one | 05:16 |
| morganfainberg | it would be a lot easier if we did config registering on import | 05:16 |
| morganfainberg | jamielennox, i'll chase it down when i talk w/ brant et al tomorrow | 05:16 |
| *** gokrokve has joined #openstack-dev | 05:17 | |
| *** morazi has joined #openstack-dev | 05:17 | |
| morganfainberg | jamielennox, in either case, i kindof think it's an auth factory... but that sjust the construct i can come up with at this point | 05:17 |
| jamielennox | yea, v2 can have a factory, v3 can have a factory | 05:17 |
| jamielennox | i can't come up with a good way to have a v2 or v3 factory | 05:17 |
| morganfainberg | auth_module.<version>_factory() | 05:18 |
| morganfainberg | it's a simple way to know if you support that API version. the method exists | 05:18 |
| *** cadenzajon has quit IRC | 05:18 | |
| *** sarob has quit IRC | 05:19 | |
| jamielennox | i meant more something that could take parameters and try v2 and v3 optimistically, there are too many possible plugins | 05:19 |
| morganfainberg | nah | 05:19 |
| morganfainberg | jamielennox, either it's a specified version, or latest stable | 05:19 |
| *** zaitcev has joined #openstack-dev | 05:19 | |
| jamielennox | morganfainberg: right but latest_stable is not so easy, there is too much difference between api versions | 05:19 |
| jamielennox | if you get a v3 when you thought latest was v2 it simply won't work | 05:20 |
| morganfainberg | jamielennox, ok ok, maybe it's latest stable _i_ support | 05:20 |
| jamielennox | i assume v4 will one day have the same problem | 05:20 |
| *** buzztroll has joined #openstack-dev | 05:20 | |
| morganfainberg | discover, get_lastest_stable([my,options,]) | 05:20 |
| jamielennox | morganfainberg: yea it's options like that i've been circling i just can't see one that feels right | 05:21 |
| *** MaxV has joined #openstack-dev | 05:21 | |
| morganfainberg | jamielennox, i think this is going to be a bite the bullet and go with a passable option | 05:22 |
| jamielennox | morganfainberg: yea, build the primitive for now, later something good might come up | 05:22 |
| morganfainberg | jamielennox, ++ | 05:22 |
| morganfainberg | jamielennox, worst case, we add a new interface that is way cooler | 05:22 |
| morganfainberg | jamielennox, deprecate the old interface in the client, tell people to move. | 05:23 |
| morganfainberg | jamielennox, or keep the old one and when everyone has moved, it becomes "internal" since it's the primitive | 05:23 |
| morganfainberg | jamielennox, but def. start small, work up | 05:23 |
| *** sarob has joined #openstack-dev | 05:23 | |
| jamielennox | morganfainberg: yep | 05:24 |
| jamielennox | morganfainberg: cheers - i need to get off the computer for a bit let my head clear | 05:24 |
| morganfainberg | jamielennox, could i purchase a +2/+A on an easy review? | 05:24 |
| morganfainberg | jamielennox, https://review.openstack.org/#/c/70257/ | 05:24 |
| morganfainberg | really easy, just cleanup of files we don't use | 05:25 |
| morganfainberg | diablo/essex | 05:25 |
| *** mriedem has quit IRC | 05:25 | |
| morganfainberg | jamielennox, :P | 05:25 |
| morganfainberg | jamielennox, if not no worries go clear your head :) | 05:25 |
| *** MaxV has quit IRC | 05:27 | |
| jamielennox | morganfainberg: we don't support the migration any longer? | 05:28 |
| morganfainberg | jamielennox, well i was also planning on smushing our migration files down | 05:28 |
| *** sarob has quit IRC | 05:28 | |
| morganfainberg | jamielennox, i don't see anything that references those files | 05:29 |
| morganfainberg | jamielennox, so not sure how they'd be used. | 05:29 |
| jamielennox | morganfainberg: although i guess it doesn't matter as you removed them with no impact on tests | 05:29 |
| morganfainberg | jamielennox, i think we stopped supporting the direct upgrade in H | 05:29 |
| jamielennox | morganfainberg: done | 05:29 |
| morganfainberg | jamielennox, cool. | 05:29 |
| jamielennox | return the favour | 05:30 |
| morganfainberg | jamielennox, once this feature proposal stuff clears up i'll see if we can collapse down the migrations | 05:30 |
| jamielennox | morganfainberg: https://review.openstack.org/#/c/70634/ | 05:30 |
| morganfainberg | jamielennox, sure | 05:30 |
| jamielennox | not quite as simple - but simple | 05:30 |
| *** ramishra_ has joined #openstack-dev | 05:30 | |
| morganfainberg | ah curl debug stuff | 05:30 |
| *** ramishra has quit IRC | 05:31 | |
| jamielennox | it was added in around the time i did the move to session and it got lost | 05:31 |
| *** mohits has joined #openstack-dev | 05:31 | |
| *** nshaikh has joined #openstack-dev | 05:31 | |
| morganfainberg | yeah | 05:31 |
| morganfainberg | looks good to me | 05:31 |
| jamielennox | yea, once it passes tests and bk's sharp eyes i figure it's mostly home | 05:32 |
| morganfainberg | jamielennox, heeh | 05:33 |
| *** otherwiseguy has quit IRC | 05:33 | |
| morganfainberg | jamielennox, this one is straightforward | 05:33 |
| morganfainberg | when i was sick it look a lot out of me | 05:33 |
| morganfainberg | it's been tough getitng back to the point where i can sit down and review code | 05:34 |
| morganfainberg | jamielennox, anyways good night dude | 05:34 |
| jamielennox | morganfainberg: thanks for that, night | 05:34 |
| *** irenab has joined #openstack-dev | 05:35 | |
| *** mgagne has quit IRC | 05:35 | |
| *** tkay has joined #openstack-dev | 05:35 | |
| *** gordc has quit IRC | 05:39 | |
| *** rohitk has joined #openstack-dev | 05:40 | |
| *** mgagne has joined #openstack-dev | 05:41 | |
| *** kobier has joined #openstack-dev | 05:41 | |
| *** garyk has quit IRC | 05:41 | |
| *** nosnos_ has joined #openstack-dev | 05:43 | |
| *** nosnos_ has quit IRC | 05:44 | |
| *** nosnos_ has joined #openstack-dev | 05:44 | |
| *** marcoemorais1 has quit IRC | 05:44 | |
| *** DinaBelova_ is now known as DinaBelova | 05:45 | |
| *** nosnos has quit IRC | 05:46 | |
| *** doug_shelley66 has joined #openstack-dev | 05:49 | |
| *** yuan has joined #openstack-dev | 05:52 | |
| *** doug_shelley66 has quit IRC | 05:53 | |
| *** jamielennox is now known as jamielennox|away | 05:55 | |
| *** yuan has quit IRC | 05:56 | |
| *** anniec has joined #openstack-dev | 06:02 | |
| *** bhuvan has joined #openstack-dev | 06:03 | |
| *** saju_m has joined #openstack-dev | 06:03 | |
| *** nosnos_ has quit IRC | 06:04 | |
| *** nosnos has joined #openstack-dev | 06:04 | |
| *** gokrokve has quit IRC | 06:04 | |
| *** arosen has joined #openstack-dev | 06:05 | |
| *** anniec_ has joined #openstack-dev | 06:05 | |
| *** anniec has quit IRC | 06:06 | |
| *** anniec_ is now known as anniec | 06:06 | |
| *** sarob has joined #openstack-dev | 06:07 | |
| *** zaitcev has quit IRC | 06:07 | |
| *** e0ne has joined #openstack-dev | 06:08 | |
| *** e0ne has quit IRC | 06:09 | |
| *** morganfainberg is now known as morganfainberg_Z | 06:12 | |
| *** jaypipes has quit IRC | 06:12 | |
| *** dkehn__ has joined #openstack-dev | 06:14 | |
| *** jaypipes has joined #openstack-dev | 06:15 | |
| *** dkehn__ has quit IRC | 06:16 | |
| *** dkehn__ has joined #openstack-dev | 06:17 | |
| *** dkehn_ has quit IRC | 06:17 | |
| *** MaxV has joined #openstack-dev | 06:24 | |
| *** ykhodork has joined #openstack-dev | 06:26 | |
| *** MaxV has quit IRC | 06:28 | |
| *** yolanda has joined #openstack-dev | 06:32 | |
| *** lifeless has quit IRC | 06:34 | |
| *** gokrokve has joined #openstack-dev | 06:35 | |
| *** irenab has quit IRC | 06:35 | |
| *** denis_makogon has joined #openstack-dev | 06:35 | |
| *** dkehn_ has joined #openstack-dev | 06:36 | |
| *** yeylon__ has quit IRC | 06:36 | |
| *** gokrokve_ has joined #openstack-dev | 06:37 | |
| *** dkehn__ has quit IRC | 06:39 | |
| *** rdas has joined #openstack-dev | 06:40 | |
| *** gokrokve has quit IRC | 06:40 | |
| *** max_lobur_afk has quit IRC | 06:40 | |
| *** aditirav has quit IRC | 06:40 | |
| *** max_lobur_afk has joined #openstack-dev | 06:41 | |
| *** neeti has joined #openstack-dev | 06:41 | |
| *** rraja has quit IRC | 06:41 | |
| *** aditirav has joined #openstack-dev | 06:41 | |
| *** diakunchikov has quit IRC | 06:41 | |
| *** diakunchikov has joined #openstack-dev | 06:41 | |
| *** anniec has quit IRC | 06:41 | |
| *** romcheg has quit IRC | 06:42 | |
| *** gokrokve_ has quit IRC | 06:42 | |
| *** markwash has quit IRC | 06:42 | |
| *** romcheg has joined #openstack-dev | 06:42 | |
| *** arosen has quit IRC | 06:43 | |
| *** dshulyak has quit IRC | 06:44 | |
| *** andreykurilin has quit IRC | 06:44 | |
| *** max_lobur_afk has quit IRC | 06:44 | |
| *** bogdando has quit IRC | 06:44 | |
| *** diakunchikov has quit IRC | 06:44 | |
| *** romcheg has quit IRC | 06:44 | |
| *** mikeoutland has joined #openstack-dev | 06:44 | |
| *** max_lobur_afk has joined #openstack-dev | 06:44 | |
| *** max_lobur_afk has quit IRC | 06:45 | |
| *** diakunchikov has joined #openstack-dev | 06:45 | |
| *** bogdando has joined #openstack-dev | 06:45 | |
| *** max_lobur_afk has joined #openstack-dev | 06:45 | |
| *** romcheg has joined #openstack-dev | 06:45 | |
| *** dshulyak has joined #openstack-dev | 06:45 | |
| *** garyk has joined #openstack-dev | 06:46 | |
| *** sarob has quit IRC | 06:47 | |
| *** sarob has joined #openstack-dev | 06:48 | |
| *** vartom1111111111 has joined #openstack-dev | 06:48 | |
| *** zoresvit1 has joined #openstack-dev | 06:51 | |
| *** zoresvit has quit IRC | 06:52 | |
| *** tdruiva has quit IRC | 06:52 | |
| *** sarob has quit IRC | 06:53 | |
| *** aditirav has quit IRC | 06:53 | |
| *** cnesa has joined #openstack-dev | 06:54 | |
| *** aditirav has joined #openstack-dev | 06:56 | |
| *** pradeep1 has joined #openstack-dev | 06:57 | |
| *** erkules_ is now known as erkules | 06:58 | |
| *** pradeep has quit IRC | 06:58 | |
| *** hdd_ has quit IRC | 06:59 | |
| *** rraja has joined #openstack-dev | 06:59 | |
| *** tkammer has joined #openstack-dev | 07:00 | |
| *** gokrokve has joined #openstack-dev | 07:00 | |
| *** eglynn has joined #openstack-dev | 07:01 | |
| *** dkehn__ has joined #openstack-dev | 07:01 | |
| *** andreykurilin has joined #openstack-dev | 07:01 | |
| *** dkehn_ has quit IRC | 07:04 | |
| *** ramishra_ has quit IRC | 07:04 | |
| *** MaxV has joined #openstack-dev | 07:04 | |
| *** nspmangalore has joined #openstack-dev | 07:05 | |
| *** gokrokve has quit IRC | 07:05 | |
| *** jtomasek has joined #openstack-dev | 07:07 | |
| nspmangalore | hi.. I'm trying to understand what the client_timeout parameter actually means | 07:09 |
| nspmangalore | Feb 12 12:30:49 ubuntuvm proxy-server: ERROR Client read timeout (60s) (txn: tx4146422d6b5e4262b8834-0052fb1bcc) (client_ip: 172.16.10.79) | 07:09 |
| *** eglynn has quit IRC | 07:09 | |
| nspmangalore | this is different from node_timeout parameter, right? | 07:09 |
| *** ramishra has joined #openstack-dev | 07:13 | |
| *** Drankis has joined #openstack-dev | 07:13 | |
| *** aswadrangnekar has joined #openstack-dev | 07:13 | |
| *** buzztroll has quit IRC | 07:14 | |
| *** buzztroll has joined #openstack-dev | 07:14 | |
| *** mkollaro has joined #openstack-dev | 07:16 | |
| *** athomas has joined #openstack-dev | 07:17 | |
| *** sarob has joined #openstack-dev | 07:18 | |
| *** rmk has joined #openstack-dev | 07:18 | |
| *** garyk has quit IRC | 07:18 | |
| *** buzztroll has quit IRC | 07:19 | |
| *** mikeoutland has quit IRC | 07:20 | |
| *** cfriesen has quit IRC | 07:20 | |
| *** coolsvap has joined #openstack-dev | 07:21 | |
| *** tkay has quit IRC | 07:21 | |
| *** aditirav has quit IRC | 07:22 | |
| *** sarob has quit IRC | 07:24 | |
| *** eglynn has joined #openstack-dev | 07:25 | |
| *** coolsvap1 has joined #openstack-dev | 07:26 | |
| *** tkay has joined #openstack-dev | 07:26 | |
| *** coolsvap has quit IRC | 07:28 | |
| *** tkay has quit IRC | 07:29 | |
| *** cnesa has quit IRC | 07:30 | |
| *** amuller has joined #openstack-dev | 07:30 | |
| *** aditirav has joined #openstack-dev | 07:30 | |
| *** eglynn has quit IRC | 07:32 | |
| *** MaxV has quit IRC | 07:33 | |
| *** tkay has joined #openstack-dev | 07:33 | |
| *** sushil_ has joined #openstack-dev | 07:33 | |
| *** jhesketh__ has quit IRC | 07:34 | |
| *** abhirc_ has quit IRC | 07:34 | |
| *** jhesketh__ has joined #openstack-dev | 07:34 | |
| *** Ursinha is now known as Ursinha-afk | 07:34 | |
| *** AlexF has joined #openstack-dev | 07:36 | |
| *** coolsvap1 has quit IRC | 07:36 | |
| *** coolsvap has joined #openstack-dev | 07:36 | |
| *** gnorth has joined #openstack-dev | 07:38 | |
| *** sushil_ has quit IRC | 07:38 | |
| *** cnesa has joined #openstack-dev | 07:38 | |
| *** sushil_ has joined #openstack-dev | 07:39 | |
| *** geraint2 has joined #openstack-dev | 07:44 | |
| *** gnorth has quit IRC | 07:45 | |
| *** corXi has joined #openstack-dev | 07:46 | |
| *** Ursinha-afk is now known as Ursinha | 07:46 | |
| *** flaper87|afk is now known as flaper87 | 07:47 | |
| *** eglynn has joined #openstack-dev | 07:47 | |
| *** e0ne has joined #openstack-dev | 07:53 | |
| *** ykhodork has quit IRC | 07:53 | |
| *** mohits has quit IRC | 07:53 | |
| *** jprovazn has joined #openstack-dev | 07:55 | |
| *** evgenyf has joined #openstack-dev | 07:55 | |
| *** rohitk has quit IRC | 07:55 | |
| *** rohitk_ has joined #openstack-dev | 07:56 | |
| *** bhuvan has quit IRC | 07:57 | |
| *** gokrokve has joined #openstack-dev | 08:00 | |
| *** marekd|away is now known as marekd | 08:03 | |
| *** rohitk_ has quit IRC | 08:03 | |
| *** gokrokve has quit IRC | 08:05 | |
| *** ykhodork has joined #openstack-dev | 08:06 | |
| *** YorikSar has quit IRC | 08:06 | |
| *** YorikSar has joined #openstack-dev | 08:06 | |
| *** amotoki has joined #openstack-dev | 08:06 | |
| *** eglynn has quit IRC | 08:07 | |
| *** YorikSar has quit IRC | 08:07 | |
| *** YorikSar has joined #openstack-dev | 08:07 | |
| *** arosen has joined #openstack-dev | 08:07 | |
| *** YorikSar has quit IRC | 08:08 | |
| *** YorikSar has joined #openstack-dev | 08:08 | |
| *** AlexF has quit IRC | 08:10 | |
| *** avishay has quit IRC | 08:10 | |
| *** tkay has quit IRC | 08:10 | |
| *** bauzas has joined #openstack-dev | 08:12 | |
| *** avishay has joined #openstack-dev | 08:12 | |
| *** nmagnezi has joined #openstack-dev | 08:12 | |
| *** jcoufal has joined #openstack-dev | 08:13 | |
| *** bhuvan has joined #openstack-dev | 08:14 | |
| *** jamieh has joined #openstack-dev | 08:15 | |
| *** devoid has quit IRC | 08:15 | |
| *** romcheg1 has joined #openstack-dev | 08:16 | |
| *** julienvey_ has joined #openstack-dev | 08:17 | |
| *** rohitk_ has joined #openstack-dev | 08:17 | |
| *** xga has joined #openstack-dev | 08:19 | |
| *** sarob has joined #openstack-dev | 08:19 | |
| *** rraja has quit IRC | 08:21 | |
| *** dkehn_ has joined #openstack-dev | 08:24 | |
| *** sarob has quit IRC | 08:24 | |
| *** zoresvit1 has quit IRC | 08:24 | |
| *** buzztroll has joined #openstack-dev | 08:25 | |
| *** dkehn_ has quit IRC | 08:26 | |
| *** dkehn_ has joined #openstack-dev | 08:26 | |
| *** dkehn__ has quit IRC | 08:27 | |
| *** yeylon__ has joined #openstack-dev | 08:27 | |
| *** mrunge has joined #openstack-dev | 08:28 | |
| *** salv-orlando has joined #openstack-dev | 08:29 | |
| *** buzztroll has quit IRC | 08:29 | |
| *** xga_ has joined #openstack-dev | 08:29 | |
| *** boris-42_ has joined #openstack-dev | 08:30 | |
| *** pschaef has joined #openstack-dev | 08:31 | |
| *** xgsa has joined #openstack-dev | 08:32 | |
| *** bhuvan has quit IRC | 08:32 | |
| *** xga has quit IRC | 08:32 | |
| *** praneshp has quit IRC | 08:33 | |
| *** eglynn has joined #openstack-dev | 08:34 | |
| *** viktors has joined #openstack-dev | 08:34 | |
| *** gszasz has joined #openstack-dev | 08:35 | |
| *** sahid has joined #openstack-dev | 08:36 | |
| *** ala has joined #openstack-dev | 08:36 | |
| *** zhikunliu has quit IRC | 08:36 | |
| *** geraint2 has quit IRC | 08:38 | |
| *** ijw_ has quit IRC | 08:41 | |
| *** nmagnezi has quit IRC | 08:42 | |
| *** jordanP has joined #openstack-dev | 08:43 | |
| *** AnilV4 has joined #openstack-dev | 08:44 | |
| *** I159 has joined #openstack-dev | 08:44 | |
| *** AnilV4 has quit IRC | 08:44 | |
| *** AnilV4 has joined #openstack-dev | 08:45 | |
| *** iartarisi has joined #openstack-dev | 08:47 | |
| *** MaxV has joined #openstack-dev | 08:49 | |
| *** nacim has joined #openstack-dev | 08:50 | |
| *** nmagnezi has joined #openstack-dev | 08:51 | |
| *** nacim has quit IRC | 08:51 | |
| *** mmagr has joined #openstack-dev | 08:51 | |
| *** nacim has joined #openstack-dev | 08:53 | |
| *** gimps has quit IRC | 08:54 | |
| Lexis | Hello everyone | 08:56 |
| Lexis | I was looking at updating OpenStack vRouter to a commercial one | 08:57 |
| Lexis | to that extend, would you recommend developing a Neutron Plugin | 08:57 |
| *** bvandenh has joined #openstack-dev | 08:58 | |
| Lexis | or update the commercial vRouter to accept calls directly from Neutron ? | 08:58 |
| *** coolsvap1 has joined #openstack-dev | 08:59 | |
| *** gokrokve has joined #openstack-dev | 09:00 | |
| *** florentflament has joined #openstack-dev | 09:00 | |
| *** coolsvap has quit IRC | 09:01 | |
| *** ifarkas has joined #openstack-dev | 09:02 | |
| *** Clabbe has quit IRC | 09:03 | |
| *** avishayb has joined #openstack-dev | 09:03 | |
| *** jistr has joined #openstack-dev | 09:04 | |
| *** jistr is now known as jistr|sick | 09:04 | |
| *** xchu has joined #openstack-dev | 09:04 | |
| *** ykhodork has quit IRC | 09:05 | |
| *** gokrokve has quit IRC | 09:05 | |
| *** ala has quit IRC | 09:05 | |
| *** ijw has joined #openstack-dev | 09:06 | |
| *** amuller has quit IRC | 09:06 | |
| *** amuller_ has joined #openstack-dev | 09:06 | |
| *** zzelle has joined #openstack-dev | 09:06 | |
| *** dtantsur has joined #openstack-dev | 09:06 | |
| *** ykhodork has joined #openstack-dev | 09:06 | |
| *** jpich has joined #openstack-dev | 09:06 | |
| *** afazekas has joined #openstack-dev | 09:07 | |
| *** gcha has joined #openstack-dev | 09:07 | |
| *** pixelb has joined #openstack-dev | 09:08 | |
| *** afazekas is now known as afazekas|sick | 09:08 | |
| *** ykhodork has quit IRC | 09:08 | |
| *** sarob has joined #openstack-dev | 09:08 | |
| *** jordanP has quit IRC | 09:09 | |
| *** ygbo has joined #openstack-dev | 09:10 | |
| *** ala has joined #openstack-dev | 09:11 | |
| *** Clabbe has joined #openstack-dev | 09:11 | |
| *** ramishra has quit IRC | 09:13 | |
| *** sarob has quit IRC | 09:13 | |
| *** lucasagomes has joined #openstack-dev | 09:15 | |
| *** jordanP has joined #openstack-dev | 09:15 | |
| *** kosi has joined #openstack-dev | 09:16 | |
| *** yassine has joined #openstack-dev | 09:18 | |
| *** sarob has joined #openstack-dev | 09:19 | |
| *** e0ne has quit IRC | 09:20 | |
| *** e0ne has joined #openstack-dev | 09:21 | |
| *** sarob has quit IRC | 09:24 | |
| *** I159_ has joined #openstack-dev | 09:24 | |
| *** safchain has joined #openstack-dev | 09:25 | |
| *** garyk has joined #openstack-dev | 09:25 | |
| *** locke105 has quit IRC | 09:26 | |
| *** rohitk_ has quit IRC | 09:29 | |
| *** rohitk has joined #openstack-dev | 09:30 | |
| *** jcoufal has quit IRC | 09:30 | |
| *** andreaf has quit IRC | 09:30 | |
| *** jcoufal has joined #openstack-dev | 09:30 | |
| *** andreaf has joined #openstack-dev | 09:31 | |
| *** athomas has quit IRC | 09:36 | |
| *** ramishra has joined #openstack-dev | 09:37 | |
| *** mkollaro has quit IRC | 09:37 | |
| *** nagyz has joined #openstack-dev | 09:38 | |
| nagyz | morning | 09:38 |
| nagyz | one of my changes got merged by jenkins into the git repo, yet I don't see myself anywhere - neither in the notes (the commit has 0), nor at the author field in the message | 09:39 |
| *** jckasper has quit IRC | 09:39 | |
| nagyz | and this seems to be the case for the majority of people | 09:39 |
| nagyz | yet, some people's commits show up properly | 09:39 |
| nagyz | am I doing something wrong? | 09:39 |
| *** DinaBelova is now known as DinaBelova_ | 09:40 | |
| *** yaguang has quit IRC | 09:40 | |
| *** Drankis has quit IRC | 09:42 | |
| jpich | nagyz: link? | 09:43 |
| gilliard | just `git log` in any OS repo | 09:43 |
| gilliard | almost all commits are by Jenkins. | 09:44 |
| nagyz | yes, but there are a lot which aren't | 09:44 |
| nagyz | how come? | 09:44 |
| nagyz | jpich, this one for example: https://github.com/openstack/nova/commit/4b7401ccc6a1677b698538e1ba4f2a515630ac67 | 09:44 |
| *** jp_at_hp has joined #openstack-dev | 09:44 | |
| nagyz | versus this one: https://github.com/openstack/nova/commit/58f5579630176c8b23c528b5c5e22b12e8354363 | 09:44 |
| gilliard | I have wondered about this too. | 09:45 |
| *** irenab has joined #openstack-dev | 09:45 | |
| gilliard | My best guess is that the ones which don't have Jenkins as an author did not need merging in the gate. I.E. the gate queue was empty when they were approved. | 09:46 |
| gilliard | I'd like to know for sure though... | 09:46 |
| nagyz | shall I open an infra bug? :-) | 09:46 |
| nagyz | but someone here must have some clue | 09:46 |
| jpich | nagyz: The original commit usually shows at the date the patch is set to (e.g. https://github.com/openstack/nova/commit/efbef496dc17adc4e4187985d13bf815d0c254a1 ) for that first example, then Jenkins also creates a merge commit if needed | 09:47 |
| nagyz | hmm | 09:48 |
| nagyz | I see | 09:48 |
| *** rraja has joined #openstack-dev | 09:48 | |
| *** amuller_ has quit IRC | 09:48 | |
| flaper87 | viktors: ping | 09:49 |
| *** qba73 has joined #openstack-dev | 09:49 | |
| *** rpodolyaka has joined #openstack-dev | 09:51 | |
| viktors | flaper87: pong | 09:52 |
| flaper87 | viktors: you're involved in oslo.db, aren't you? | 09:54 |
| viktors | flaper87: yes, you right | 09:56 |
| *** sushil_ has quit IRC | 09:56 | |
| *** sushil_ has joined #openstack-dev | 09:57 | |
| *** denis_makogon has quit IRC | 09:57 | |
| *** dperaza has joined #openstack-dev | 10:00 | |
| *** gokrokve has joined #openstack-dev | 10:00 | |
| *** sushil_ has quit IRC | 10:02 | |
| *** zzelle has quit IRC | 10:02 | |
| *** Steap_ has quit IRC | 10:03 | |
| *** dperaza has quit IRC | 10:04 | |
| *** Steap_ has joined #openstack-dev | 10:04 | |
| *** gokrokve has quit IRC | 10:05 | |
| *** ijw has quit IRC | 10:06 | |
| *** sweston has quit IRC | 10:08 | |
| *** rohitk_ has joined #openstack-dev | 10:09 | |
| *** markmc has joined #openstack-dev | 10:10 | |
| *** rohitk has quit IRC | 10:11 | |
| *** MaxV has quit IRC | 10:11 | |
| *** MaxV has joined #openstack-dev | 10:12 | |
| *** MaxV has quit IRC | 10:12 | |
| *** rossella_s has joined #openstack-dev | 10:12 | |
| *** ramishra has quit IRC | 10:14 | |
| *** sushil_ has joined #openstack-dev | 10:14 | |
| *** xga_ has quit IRC | 10:15 | |
| *** MaxV has joined #openstack-dev | 10:15 | |
| *** xga has joined #openstack-dev | 10:18 | |
| *** CaptTofu has joined #openstack-dev | 10:19 | |
| *** gszasz has quit IRC | 10:19 | |
| *** sarob has joined #openstack-dev | 10:19 | |
| *** MaxV has quit IRC | 10:20 | |
| *** rohitk_ is now known as rohitk | 10:21 | |
| *** xchu has quit IRC | 10:22 | |
| *** MaxV has joined #openstack-dev | 10:22 | |
| *** xqueralt has joined #openstack-dev | 10:24 | |
| *** danpb has joined #openstack-dev | 10:24 | |
| *** sarob has quit IRC | 10:24 | |
| *** xuhanp has quit IRC | 10:28 | |
| *** paragan has quit IRC | 10:29 | |
| *** MaxV has quit IRC | 10:31 | |
| *** kolesovdv has joined #openstack-dev | 10:32 | |
| *** schwicke has joined #openstack-dev | 10:32 | |
| *** schwicke has joined #openstack-dev | 10:33 | |
| *** schwicke has quit IRC | 10:33 | |
| *** martyntaylor has joined #openstack-dev | 10:33 | |
| *** martyntaylor has quit IRC | 10:33 | |
| *** CaptTofu has quit IRC | 10:40 | |
| *** amuller_ has joined #openstack-dev | 10:41 | |
| *** matsuhashi has quit IRC | 10:41 | |
| *** markmc has quit IRC | 10:43 | |
| *** gongysh has quit IRC | 10:44 | |
| *** evgenyf has quit IRC | 10:44 | |
| *** VINOD has joined #openstack-dev | 10:45 | |
| *** vartom1111111112 has joined #openstack-dev | 10:46 | |
| *** lbragstad_ has quit IRC | 10:46 | |
| *** rodrigods has joined #openstack-dev | 10:46 | |
| *** rodrigods has joined #openstack-dev | 10:46 | |
| *** lbragstad has quit IRC | 10:46 | |
| *** amuller_ has quit IRC | 10:47 | |
| *** markmc has joined #openstack-dev | 10:48 | |
| *** vartom1111111111 has quit IRC | 10:48 | |
| *** lbragstad_ has joined #openstack-dev | 10:49 | |
| *** lbragstad has joined #openstack-dev | 10:50 | |
| *** VINOD has quit IRC | 10:50 | |
| *** martyntaylor has joined #openstack-dev | 10:52 | |
| *** markmc has quit IRC | 10:53 | |
| *** evgenyf has joined #openstack-dev | 10:54 | |
| *** markmc has joined #openstack-dev | 10:57 | |
| *** saju_m has quit IRC | 10:57 | |
| *** jistr|sick has quit IRC | 10:57 | |
| *** gokrokve has joined #openstack-dev | 11:00 | |
| *** max_lobur_afk is now known as max_lobur | 11:03 | |
| *** CaptTofu has joined #openstack-dev | 11:04 | |
| *** gokrokve has quit IRC | 11:05 | |
| *** markmc has quit IRC | 11:06 | |
| *** samuelbercovici has joined #openstack-dev | 11:07 | |
| *** gszasz has joined #openstack-dev | 11:07 | |
| *** lbragstad has quit IRC | 11:07 | |
| *** lifeless has joined #openstack-dev | 11:07 | |
| *** lbragstad_ has quit IRC | 11:07 | |
| *** henrynash has joined #openstack-dev | 11:08 | |
| *** amcrn has quit IRC | 11:08 | |
| *** lbragstad_ has joined #openstack-dev | 11:09 | |
| *** lbragstad has joined #openstack-dev | 11:10 | |
| *** markmc has joined #openstack-dev | 11:12 | |
| *** YorikSar has quit IRC | 11:14 | |
| *** nmagnezi has quit IRC | 11:15 | |
| *** xga has quit IRC | 11:15 | |
| *** dperaza has joined #openstack-dev | 11:18 | |
| *** dkehn__ has joined #openstack-dev | 11:18 | |
| *** sarob has joined #openstack-dev | 11:19 | |
| *** amuller_ has joined #openstack-dev | 11:20 | |
| *** dkehn__ has quit IRC | 11:20 | |
| *** dkehn__ has joined #openstack-dev | 11:20 | |
| *** dkehn_ has quit IRC | 11:21 | |
| *** gcha has quit IRC | 11:21 | |
| *** amuller_ is now known as amuller | 11:22 | |
| *** rohitk has quit IRC | 11:23 | |
| *** bhuvan has joined #openstack-dev | 11:23 | |
| *** ArxCruz has joined #openstack-dev | 11:23 | |
| *** markmc has quit IRC | 11:23 | |
| *** sarob has quit IRC | 11:24 | |
| *** jruzicka has joined #openstack-dev | 11:24 | |
| *** YorikSar has joined #openstack-dev | 11:24 | |
| *** sahid has quit IRC | 11:24 | |
| *** kmartin has quit IRC | 11:26 | |
| *** branen has quit IRC | 11:26 | |
| *** markmc has joined #openstack-dev | 11:26 | |
| *** sahid has joined #openstack-dev | 11:27 | |
| *** YorikSar has quit IRC | 11:27 | |
| *** rohitk has joined #openstack-dev | 11:27 | |
| *** ala has quit IRC | 11:28 | |
| *** nebulta has joined #openstack-dev | 11:29 | |
| *** mdenny has quit IRC | 11:30 | |
| *** ilyashakhat has joined #openstack-dev | 11:30 | |
| *** amuller has quit IRC | 11:30 | |
| *** YorikSar has joined #openstack-dev | 11:30 | |
| *** YorikSar has quit IRC | 11:31 | |
| *** kosi has quit IRC | 11:32 | |
| *** YorikSar has joined #openstack-dev | 11:32 | |
| *** YorikSar has quit IRC | 11:33 | |
| *** YorikSar has joined #openstack-dev | 11:34 | |
| *** YorikSar has quit IRC | 11:34 | |
| *** bhuvan has quit IRC | 11:34 | |
| *** YorikSar has joined #openstack-dev | 11:34 | |
| *** YorikSar has quit IRC | 11:35 | |
| *** YorikSar has joined #openstack-dev | 11:35 | |
| *** markmc has quit IRC | 11:36 | |
| *** aloga has quit IRC | 11:37 | |
| *** schwicht has joined #openstack-dev | 11:37 | |
| *** saju_m has joined #openstack-dev | 11:40 | |
| *** ala has joined #openstack-dev | 11:40 | |
| *** markmc has joined #openstack-dev | 11:42 | |
| *** pcm_ has joined #openstack-dev | 11:42 | |
| *** matsuhashi has joined #openstack-dev | 11:44 | |
| *** topol has joined #openstack-dev | 11:44 | |
| *** pcm_ has quit IRC | 11:44 | |
| *** pcm_ has joined #openstack-dev | 11:45 | |
| *** corXi has quit IRC | 11:47 | |
| *** YorikSar has quit IRC | 11:48 | |
| *** saju_m has quit IRC | 11:49 | |
| *** nmagnezi has joined #openstack-dev | 11:49 | |
| *** markmc has quit IRC | 11:50 | |
| *** kosi has joined #openstack-dev | 11:50 | |
| *** YorikSar has joined #openstack-dev | 11:51 | |
| *** YorikSar has quit IRC | 11:51 | |
| *** YorikSar has joined #openstack-dev | 11:52 | |
| *** pschaef has quit IRC | 11:52 | |
| *** b3nt_pin has joined #openstack-dev | 11:53 | |
| *** bhuvan has joined #openstack-dev | 11:54 | |
| *** kosi has quit IRC | 11:54 | |
| *** markmc has joined #openstack-dev | 11:55 | |
| *** CaptTofu has quit IRC | 11:55 | |
| *** pschaef has joined #openstack-dev | 11:56 | |
| *** julienvey_ has quit IRC | 11:58 | |
| *** kosi has joined #openstack-dev | 11:58 | |
| *** gokrokve has joined #openstack-dev | 12:00 | |
| *** doug_shelley66 has joined #openstack-dev | 12:00 | |
| *** kosi has quit IRC | 12:00 | |
| *** MaxV has joined #openstack-dev | 12:01 | |
| *** rraja has quit IRC | 12:01 | |
| *** DinaBelova_ is now known as DinaBelova | 12:04 | |
| *** julienvey_ has joined #openstack-dev | 12:04 | |
| *** gokrokve has quit IRC | 12:05 | |
| *** amuller has joined #openstack-dev | 12:05 | |
| *** rfolco has joined #openstack-dev | 12:08 | |
| *** julienvey_ has quit IRC | 12:09 | |
| *** jhesketh__ has quit IRC | 12:12 | |
| *** coolsvap_ has joined #openstack-dev | 12:14 | |
| *** samuelbercovici1 has joined #openstack-dev | 12:15 | |
| *** xBsd has joined #openstack-dev | 12:15 | |
| *** samuelbercovici has quit IRC | 12:17 | |
| *** samuelbercovici1 is now known as samuelbercovici | 12:17 | |
| *** coolsvap1 has quit IRC | 12:17 | |
| *** b3nt_pin has quit IRC | 12:17 | |
| *** dkehn_ has joined #openstack-dev | 12:18 | |
| *** julienvey_ has joined #openstack-dev | 12:18 | |
| *** kosi has joined #openstack-dev | 12:18 | |
| *** sarob has joined #openstack-dev | 12:19 | |
| *** aditirav_ has joined #openstack-dev | 12:21 | |
| *** aditirav has quit IRC | 12:21 | |
| *** aditirav_ is now known as aditirav | 12:21 | |
| *** dkehn__ has quit IRC | 12:21 | |
| *** baoli has joined #openstack-dev | 12:23 | |
| *** kosi has quit IRC | 12:23 | |
| *** rohitk has quit IRC | 12:23 | |
| *** baoli has quit IRC | 12:23 | |
| *** sarob has quit IRC | 12:24 | |
| *** baoli has joined #openstack-dev | 12:25 | |
| *** paragan has joined #openstack-dev | 12:25 | |
| *** gcha has joined #openstack-dev | 12:26 | |
| *** xga has joined #openstack-dev | 12:28 | |
| *** raildo has joined #openstack-dev | 12:28 | |
| *** Sumeniac has quit IRC | 12:31 | |
| *** sgordon has joined #openstack-dev | 12:31 | |
| *** sgordon has joined #openstack-dev | 12:31 | |
| *** dperaza has quit IRC | 12:31 | |
| *** IanGovett has joined #openstack-dev | 12:32 | |
| *** YorikSar has quit IRC | 12:32 | |
| *** kosi has joined #openstack-dev | 12:32 | |
| *** YorikSar has joined #openstack-dev | 12:33 | |
| *** eglynn has quit IRC | 12:33 | |
| *** aditirav has quit IRC | 12:33 | |
| *** eharney has quit IRC | 12:36 | |
| *** FunnyLookinHat has joined #openstack-dev | 12:37 | |
| *** Tross has quit IRC | 12:38 | |
| *** rohitk has joined #openstack-dev | 12:39 | |
| *** alex-foo has joined #openstack-dev | 12:39 | |
| alex-foo | some of the jenkins tests for git-review are sporadically hanging and I can't get my patch to verify | 12:39 |
| alex-foo | https://jenkins02.openstack.org/job/gate-git-review-python26/70/ | 12:39 |
| alex-foo | https://review.openstack.org/#/c/72751 | 12:39 |
| alex-foo | also why doesn't jenkins post the job URL to gerrit? | 12:40 |
| *** lucasagomes is now known as lucas-hungry | 12:41 | |
| YorikSar | alex-foo: You have PEP8 failing. It is the only check that is stable in git-review right now. | 12:43 |
| marekd | YorikSar: BTW do you know what's going on with thegerrit? | 12:44 |
| *** Alexei_987 has joined #openstack-dev | 12:45 | |
| YorikSar | marekd: I doubt it. What's wrong with it? | 12:45 |
| *** mrunge has quit IRC | 12:45 | |
| *** doug_shelley66 has quit IRC | 12:45 | |
| marekd | YorikSar: by saying "it's the only check that is stable in ...[]" did you mean alex-foo patch and his set of tests or gerrit infra in general? | 12:46 |
| alex-foo | YorikSar: i fixed the PEP8 in Patch Set 2 | 12:46 |
| YorikSar | marekd: I mean that it's the only check that fails if and only if patch actually fails. | 12:47 |
| *** bhuvan has quit IRC | 12:47 | |
| marekd | YorikSar: okay then, i thought you were talking about the gerrit in general :-) Sorry for the mess. | 12:47 |
| YorikSar | marekd: nah... Gerrit is fine, git-review is fine. Tests for git-review require about 3 patches to become at least more stable. | 12:48 |
| alex-foo | YorikSar: just looking at the timeline it seems it took 40 minutes for the other real tests to time out | 12:48 |
| YorikSar | alex-foo: Yeah... And that's why I want to push https://review.openstack.org/71223 | 12:49 |
| alex-foo | bleh | 12:49 |
| *** dims has joined #openstack-dev | 12:50 | |
| *** aloga has joined #openstack-dev | 12:50 | |
| alex-foo | 'recheck no bug' is some special comment that jenkins uses? | 12:50 |
| *** lucas-hungry has quit IRC | 12:50 | |
| *** Sumeniac has joined #openstack-dev | 12:50 | |
| *** thouveng has quit IRC | 12:50 | |
| *** florentflament has quit IRC | 12:51 | |
| YorikSar | alex-foo: It signals Zuul to restart checks on this patch | 12:53 |
| erlon | hi guys, Im trying to bring a an openstack install with devstack but the script is stucking with a error: | 12:53 |
| erlon | 2014-02-12 07:49:31 [ERROR] /home/steve/devstack/lib/cinder:184 No suitable rootwrap found | 12:53 |
| alex-foo | YorikSar: if i can see a test is hanging can i use it to nudge it into restarting? | 12:53 |
| erlon | any idea?? | 12:53 |
| *** kenperkins has quit IRC | 12:53 | |
| YorikSar | alex-foo: You cannot abort test. But you can restart it once it finishes... | 12:54 |
| alex-foo | YorikSar: so i have to wait 40 minutes before i can iterate this test? | 12:54 |
| alex-foo | *patch | 12:54 |
| *** b3nt_pin has joined #openstack-dev | 12:55 | |
| YorikSar | alex-foo: yep... | 12:55 |
| YorikSar | alex-foo: Or you can rebase on top of my patch | 12:55 |
| *** b3nt_pin is now known as beagles | 12:55 | |
| YorikSar | alex-foo: But please don't rebase my patch acidentally (i.e. use "git review -R") | 12:56 |
| *** nspmangalore has quit IRC | 12:56 | |
| alex-foo | who has submit rights on git review? | 12:57 |
| *** nacim has quit IRC | 12:57 | |
| *** xga_ has joined #openstack-dev | 12:58 | |
| *** schwicht has quit IRC | 12:58 | |
| *** bhuvan has joined #openstack-dev | 12:58 | |
| *** xgsa has quit IRC | 12:58 | |
| *** eglynn has joined #openstack-dev | 12:58 | |
| *** jgallard has joined #openstack-dev | 12:59 | |
| YorikSar | alex-foo: I'm currently working with one of core reviewers. | 13:00 |
| *** mgoddard has quit IRC | 13:00 | |
| *** gokrokve has joined #openstack-dev | 13:00 | |
| YorikSar | alex-foo: You can find list of cores of any projects in Gerrit in Admin/Groups | 13:00 |
| *** vartom1111111113 has joined #openstack-dev | 13:00 | |
| *** henrynash has quit IRC | 13:00 | |
| *** tmclaugh[work] has joined #openstack-dev | 13:00 | |
| *** xgsa has joined #openstack-dev | 13:01 | |
| *** vartom1111111112 has quit IRC | 13:01 | |
| *** xga has quit IRC | 13:01 | |
| *** xga_ has quit IRC | 13:03 | |
| *** noslzzp has joined #openstack-dev | 13:03 | |
| *** lucas-hungry has joined #openstack-dev | 13:03 | |
| *** xga has joined #openstack-dev | 13:03 | |
| *** gokrokve has quit IRC | 13:05 | |
| *** tmclaugh[work] has quit IRC | 13:05 | |
| *** CaptTofu has joined #openstack-dev | 13:05 | |
| *** saju_m has joined #openstack-dev | 13:05 | |
| *** saju_m has quit IRC | 13:06 | |
| *** saju_m has joined #openstack-dev | 13:07 | |
| *** markvoelker1 has joined #openstack-dev | 13:07 | |
| *** Tross has joined #openstack-dev | 13:09 | |
| *** henrynash has joined #openstack-dev | 13:10 | |
| *** jecarey has quit IRC | 13:10 | |
| *** tmclaugh[work] has joined #openstack-dev | 13:10 | |
| *** Tross has quit IRC | 13:10 | |
| *** kosi has left #openstack-dev | 13:10 | |
| *** mwagner_lap has quit IRC | 13:11 | |
| *** nacim has joined #openstack-dev | 13:12 | |
| *** jay-lau-513 has quit IRC | 13:14 | |
| *** jay-lau-513 has joined #openstack-dev | 13:14 | |
| *** mgoddard has joined #openstack-dev | 13:15 | |
| alex-foo | YorikSar: i'm on my third run and python 3.3 is now hanging... so frustrating! | 13:16 |
| *** dkranz has quit IRC | 13:16 | |
| YorikSar | alex-foo: Yeah, I've seen it. | 13:16 |
| alex-foo | why are they hanging? | 13:16 |
| YorikSar | alex-foo: You can rebase your change on top of mine. They won't hang for long then. They'll fail faster. | 13:17 |
| alex-foo | YorikSar: is your change likely to be accepted soon? | 13:17 |
| YorikSar | alex-foo: I really hope so. | 13:17 |
| *** heyongli has joined #openstack-dev | 13:18 | |
| *** xgsa has quit IRC | 13:18 | |
| *** henrynash has quit IRC | 13:18 | |
| *** lbragstad has quit IRC | 13:19 | |
| *** sarob has joined #openstack-dev | 13:19 | |
| *** mkollaro has joined #openstack-dev | 13:19 | |
| *** yamahata has quit IRC | 13:19 | |
| *** xgsa has joined #openstack-dev | 13:20 | |
| *** dstanek has joined #openstack-dev | 13:20 | |
| *** galstrom_zzz is now known as galstrom | 13:21 | |
| *** eharney has joined #openstack-dev | 13:22 | |
| *** jecarey has joined #openstack-dev | 13:23 | |
| *** sarob has quit IRC | 13:24 | |
| *** henrynash has joined #openstack-dev | 13:24 | |
| *** yamahata has joined #openstack-dev | 13:25 | |
| alex-foo | interested to know how the jenkins jobs are secured against arbitrary network access etc. - any pointers? | 13:25 |
| *** pico-pete has joined #openstack-dev | 13:27 | |
| *** neeti has quit IRC | 13:28 | |
| *** jdob has joined #openstack-dev | 13:28 | |
| *** prad_ has joined #openstack-dev | 13:29 | |
| *** galstrom is now known as galstrom_zzz | 13:29 | |
| *** jhesketh_ has quit IRC | 13:30 | |
| *** henrynash has quit IRC | 13:30 | |
| *** thomasem has joined #openstack-dev | 13:31 | |
| *** galstrom_zzz is now known as galstrom | 13:31 | |
| *** thomasem has left #openstack-dev | 13:31 | |
| *** CaptTofu has quit IRC | 13:32 | |
| *** dprince has joined #openstack-dev | 13:32 | |
| *** thomasem_ has joined #openstack-dev | 13:32 | |
| *** galstrom is now known as galstrom_zzz | 13:32 | |
| *** vladikr has joined #openstack-dev | 13:32 | |
| *** kevinconway has joined #openstack-dev | 13:33 | |
| *** rtheis has joined #openstack-dev | 13:34 | |
| *** doug_shelley66 has joined #openstack-dev | 13:34 | |
| *** thouveng has joined #openstack-dev | 13:35 | |
| *** radez_g0n3 is now known as radez | 13:38 | |
| *** xBsd has quit IRC | 13:39 | |
| *** CaptTofu has joined #openstack-dev | 13:40 | |
| *** julienvey_ has quit IRC | 13:45 | |
| *** yaguang has joined #openstack-dev | 13:46 | |
| *** mfer has joined #openstack-dev | 13:46 | |
| *** julienvey_ has joined #openstack-dev | 13:47 | |
| *** mkollaro has quit IRC | 13:47 | |
| *** henrynash has joined #openstack-dev | 13:48 | |
| *** Tross has joined #openstack-dev | 13:49 | |
| *** lbragstad has joined #openstack-dev | 13:52 | |
| *** rohitk has quit IRC | 13:53 | |
| *** jprovazn has quit IRC | 13:55 | |
| *** jprovazn has joined #openstack-dev | 13:55 | |
| *** lucas-hungry is now known as lucasagomes | 13:56 | |
| *** krtaylor has quit IRC | 13:56 | |
| *** krtaylor has joined #openstack-dev | 13:57 | |
| *** Tross has quit IRC | 13:57 | |
| *** hdd_ has joined #openstack-dev | 13:58 | |
| *** YorikSar_ has joined #openstack-dev | 13:58 | |
| *** mrodden has quit IRC | 13:59 | |
| *** locke105 has joined #openstack-dev | 14:00 | |
| *** gokrokve has joined #openstack-dev | 14:00 | |
| *** heyongli has quit IRC | 14:00 | |
| *** achampion has quit IRC | 14:01 | |
| *** krtaylor has quit IRC | 14:01 | |
| YorikSar_ | fungi: Hello. Around? | 14:01 |
| *** krtaylor has joined #openstack-dev | 14:02 | |
| *** nmagnezi has quit IRC | 14:02 | |
| *** coolsvap_ has quit IRC | 14:02 | |
| *** tdruiva has joined #openstack-dev | 14:02 | |
| *** YorikSar has quit IRC | 14:03 | |
| *** YorikSar_ is now known as YorikSar | 14:03 | |
| *** sahid has quit IRC | 14:04 | |
| *** sahid has joined #openstack-dev | 14:04 | |
| *** gokrokve has quit IRC | 14:05 | |
| *** yamahata has quit IRC | 14:05 | |
| *** vartom1111111113 has quit IRC | 14:05 | |
| *** tdruiva has quit IRC | 14:05 | |
| *** zzelle has joined #openstack-dev | 14:06 | |
| *** vartom1111111113 has joined #openstack-dev | 14:06 | |
| dims | saschpe_, ping | 14:06 |
| saschpe_ | dims: pong | 14:06 |
| dims | saschpe_, need your help to unblock this review https://review.openstack.org/#/c/72330/ which is stuck because of a problem with installing data files to /etc/ i've added a rollback review https://review.openstack.org/#/c/72949/ please take a look | 14:07 |
| *** glenng has quit IRC | 14:07 | |
| *** CaptTofu has quit IRC | 14:08 | |
| fungi | YorikSar: yeah, i'm around. what do you need? | 14:08 |
| YorikSar | fungi: I want to try to push 3 changes to git-review that should make test accuracy bearable | 14:09 |
| *** tdruiva has joined #openstack-dev | 14:09 | |
| *** tdruiva has quit IRC | 14:09 | |
| *** CaptTofu has joined #openstack-dev | 14:10 | |
| saschpe_ | dims: can you point me to the gate code that installs deps? | 14:10 |
| *** glenng has joined #openstack-dev | 14:10 | |
| *** sweston has joined #openstack-dev | 14:10 | |
| *** tdruiva has joined #openstack-dev | 14:10 | |
| *** yamahata has joined #openstack-dev | 14:10 | |
| *** thuc has joined #openstack-dev | 14:10 | |
| YorikSar | fungi: SergeyLukjanov was looking at https://review.openstack.org/71223 and wants 2 more successes from Jenkins, but I'd suggest pushing it as is. It should at least make Jenkins fail fast. | 14:10 |
| *** vartom1111111113 has quit IRC | 14:10 | |
| YorikSar | fungi: https://review.openstack.org/70685 should fix about half of problems that still appear | 14:11 |
| *** nmagnezi has joined #openstack-dev | 14:12 | |
| *** cagrev has joined #openstack-dev | 14:12 | |
| YorikSar | fungi: and https://review.openstack.org/72910 should provide better logs if something fail so that we can continue investigating what's going wrong. | 14:12 |
| *** vkmc has joined #openstack-dev | 14:12 | |
| *** mrodden has joined #openstack-dev | 14:12 | |
| *** Tross has joined #openstack-dev | 14:13 | |
| fungi | YorikSar: the test completion times on all the rechecks of 71223 are very promising | 14:13 |
| YorikSar | fungi: Yeah... But it looks like we don't see proper Timeout exception because of my addition about known_hosts. | 14:14 |
| *** sweston has quit IRC | 14:14 | |
| fungi | YorikSar: i remember discussing 70685 but hadn't reviewed that latest solution. doing so now | 14:15 |
| YorikSar | fungi: It's a really great suggestion provided by cboylan :) | 14:16 |
| dims | saschpe_, no clue. looking | 14:17 |
| *** dkranz has joined #openstack-dev | 14:17 | |
| saschpe_ | dims: I suspect a pbr issue, I'll look | 14:17 |
| dims | saschpe_, thanks. | 14:17 |
| YorikSar | Oh, it's clarkb, not cboylan :) | 14:18 |
| fungi | YorikSar: yeah, he and i were network engineers in former lives, so we batted the idea of overloading the loopback /8 around for a while | 14:18 |
| *** pschaef has quit IRC | 14:18 | |
| *** doug-fish has quit IRC | 14:18 | |
| fungi | i agree it's an elegant workaround for the problem | 14:18 |
| *** xga has quit IRC | 14:18 | |
| YorikSar | clarkb: Hi, can you join us? | 14:18 |
| *** yamahata has quit IRC | 14:19 | |
| YorikSar | fungi: I'd say it's a solution, not workaround. Current implementation limits us with 1k individual tests. | 14:19 |
| fungi | YorikSar: he likely can't. the sun's not up in his timezone yet (he's further west than i am, and i only just started eating breakfast) | 14:19 |
| *** sarob has joined #openstack-dev | 14:19 | |
| YorikSar | fungi: And it can be easily scaled up to 1m by mediating ports as well. | 14:20 |
| *** hartsocks has joined #openstack-dev | 14:20 | |
| fungi | agreed. i meant it's a workaround to the problem of being unable to have gerrit grab an ephemeral port allocation from the kernel (because this is actually an integration test and we don't have low-level control over its port-binding method as a result) | 14:21 |
| *** saju_m has quit IRC | 14:21 | |
| *** rossella_s has quit IRC | 14:22 | |
| *** yamahata has joined #openstack-dev | 14:23 | |
| *** sarob has quit IRC | 14:24 | |
| fungi | YorikSar: i definitely want to merge 70685 as soon as possible, but i'm taking the opportunity of tests being broken to exercise the other two, since we won't be able to see as easily if they're doind what we want once tests consistently succeed | 14:25 |
| YorikSar | fungi: Unfortunatelly that change won't fix all tests. | 14:26 |
| YorikSar | fungi: It still failed couple times by itself. | 14:26 |
| *** joesavak has joined #openstack-dev | 14:27 | |
| fungi | patchset 3 of 70685 doesn't seem to have failed yet | 14:27 |
| *** tongli has joined #openstack-dev | 14:27 | |
| fungi | but i'm rechecking it a few more times to see what else we have going on there | 14:28 |
| *** gokrokve has joined #openstack-dev | 14:28 | |
| YorikSar | fungi: Oh, nice... | 14:28 |
| *** ayoung has joined #openstack-dev | 14:29 | |
| YorikSar | fungi: But I still wonder why would tests timeout. | 14:29 |
| YorikSar | fungi: I think, there's something wrong with Gerrit's gerrit.sh. It waits indefinitelly for Gerrit to get up. | 14:30 |
| fungi | YorikSar: anyway, some time today i'll get clarkb or SergeyLukjanov to look over 71223 and 72910 again (i've already +2'd all three of those patches) and try to get them all in | 14:30 |
| *** thuc has quit IRC | 14:30 | |
| *** Tross has quit IRC | 14:30 | |
| *** bswartz1 has joined #openstack-dev | 14:30 | |
| fungi | YorikSar: we could certainly add a timeout command in that shell script to fail it after a sane amount of time | 14:30 |
| *** thuc has joined #openstack-dev | 14:30 | |
| *** bswartz has quit IRC | 14:30 | |
| YorikSar | fungi: Yeah... But it sounds like duplication of Timeout fixture's work. | 14:31 |
| SergeyLukjanov | fungi, YorikSar, looking on them atm | 14:31 |
| BobBall | fungi: Any chance you can set up the service account for citrix_xenserver_ci? We're tyring to tie the configuration together ATM and it's a bit of a missing link :) | 14:31 |
| fungi | or if the daemon disassociates too soon from the controlling terminal, a loop and kill | 14:31 |
| *** irenab has quit IRC | 14:31 | |
| fungi | BobBall: sure thing--did you request it on the -infra ml or via bug to openstack-ci? i'm a couple days behind on ml reading and bug triage i'm afraid | 14:32 |
| BobBall | ML | 14:32 |
| BobBall | and no problem - I know things are hectic and normally we'd be quite happy to wait, but we've got the I-3 deadline looming for XenAPI :) | 14:32 |
| fungi | BobBall: awesome. i'll hunt it down and try to get through new account requests on the whole today. sorry about any delay | 14:32 |
| BobBall | Perfect, thanks! Appreciated! | 14:33 |
| fungi | assuming my electricity stays on, anyway. the national weather service is all doom and gloom for my area | 14:33 |
| BobBall | yeah - its' weird all over the place ATM! | 14:34 |
| *** gordc has joined #openstack-dev | 14:34 | |
| YorikSar | BobBall, fungi: where're you from, guys? | 14:34 |
| *** henrynash has quit IRC | 14:34 | |
| SergeyLukjanov | fungi, YorikSar, https://review.openstack.org/#/c/71223/ works again, so, I'm agreed with merging it to make tests failing faster | 14:35 |
| *** thuc has quit IRC | 14:35 | |
| BobBall | I'm UK - we've had very heavy storms here - but I've heard about the fun weather stateside too | 14:35 |
| *** cnesa has quit IRC | 14:37 | |
| *** mriedem has joined #openstack-dev | 14:37 | |
| SergeyLukjanov | fungi, YorikSar, IMO https://review.openstack.org/#/c/72910/ could be approved after at least one +1 from jenkins :) | 14:37 |
| YorikSar | SergeyLukjanov: I think it'll pass once 70685 will land :) | 14:38 |
| *** henrynash has joined #openstack-dev | 14:38 | |
| mflobo_ | Someone to code-review this patch, please? https://review.openstack.org/#/c/68603/ | 14:38 |
| *** saju_m has joined #openstack-dev | 14:38 | |
| *** mflobo_ has quit IRC | 14:38 | |
| YorikSar | But it'll take a lot of time before it fails w/o 71233... | 14:39 |
| *** mflobo has joined #openstack-dev | 14:39 | |
| SergeyLukjanov | YorikSar, 70865 approved due 3x +2 | 14:39 |
| *** clayb has joined #openstack-dev | 14:39 | |
| SergeyLukjanov | fungi, are you ok with approving 71233? it looks like that it at least help us to push other fixes | 14:40 |
| YorikSar | SergeyLukjanov: Oh... fungi wanted to keep it hanging to let others to be checked with failing gate... | 14:40 |
| *** jmckind has joined #openstack-dev | 14:40 | |
| *** aditirav has joined #openstack-dev | 14:40 | |
| YorikSar | SergeyLukjanov: I mean, 70865 | 14:40 |
| *** rossella_s has joined #openstack-dev | 14:41 | |
| *** bswartz1 has quit IRC | 14:42 | |
| *** pschaef has joined #openstack-dev | 14:42 | |
| SergeyLukjanov | YorikSar, heh, I've -2'd it, can't remember actually will it stop zuul or not ;) | 14:42 |
| *** mgagne has quit IRC | 14:42 | |
| *** bswartz has joined #openstack-dev | 14:44 | |
| *** mgagne has joined #openstack-dev | 14:44 | |
| fungi | YorikSar: i'm in the middle of the eastern coast of the usa. we have significant ice storms predicted for later today | 14:44 |
| YorikSar | SergeyLukjanov: Looks like it's going to hang anyway... | 14:44 |
| *** mgagne has quit IRC | 14:44 | |
| *** aeperezt has joined #openstack-dev | 14:44 | |
| *** galstrom_zzz is now known as galstrom | 14:44 | |
| *** cnesa has joined #openstack-dev | 14:45 | |
| SergeyLukjanov | YorikSar, yup | 14:45 |
| fungi | YorikSar: SergeyLukjanov: we've probably already collected enough data on the multiple rechecks of the other two changes at this point. we can always emulate timeouts with a throwaway patch later if we want | 14:45 |
| *** armax has joined #openstack-dev | 14:45 | |
| YorikSar | Good news: 70685 won't cure us entirely :) | 14:45 |
| fungi | heh | 14:45 |
| *** cfriesen has joined #openstack-dev | 14:46 | |
| SergeyLukjanov | YorikSar, bah | 14:46 |
| *** jamieh has quit IRC | 14:47 | |
| *** mohits has joined #openstack-dev | 14:47 | |
| *** jamieh has joined #openstack-dev | 14:47 | |
| *** peristeri has joined #openstack-dev | 14:48 | |
| SergeyLukjanov | so, we have a patch with 1) 2m timeout to prevent 1h hangs for test, 2) gerrit port selection improvements and 3) known_hosts attachment fix | 14:48 |
| SergeyLukjanov | #1 and #3 could help us to debug/land other fix | 14:48 |
| *** mkollaro has joined #openstack-dev | 14:48 | |
| SergeyLukjanov | fungi, YorikSar, ^^ | 14:48 |
| YorikSar | SergeyLukjanov, fungi: how about we push 2m timeout change? It looks like the other two will hang for at least 40m now. | 14:48 |
| saschpe_ | dims: revert was merged, that should unblock it. i have to discuss the underlying setuptools issue upstream first before reattempting this :-) | 14:49 |
| *** bhuvan has quit IRC | 14:49 | |
| SergeyLukjanov | YorikSar, I'm ok with landing timeout and known_hosts attachment fix | 14:49 |
| YorikSar | At least I'll be able to reverify them during my evening :) | 14:49 |
| saschpe_ | in short, --prefix is added in front of all data_files, which is wrong | 14:49 |
| *** nermina has joined #openstack-dev | 14:50 | |
| fungi | YorikSar: SergeyLukjanov: sounds great | 14:50 |
| *** stevemar has joined #openstack-dev | 14:50 | |
| *** mriedem has quit IRC | 14:51 | |
| SergeyLukjanov | fungi, YorikSar, approving them and let's look on the port selection | 14:51 |
| *** pschaef has quit IRC | 14:51 | |
| *** mohits has quit IRC | 14:52 | |
| SergeyLukjanov | fungi, do you remember, will zull reject change if it'll be -2'd after approving? | 14:52 |
| *** cnesa has quit IRC | 14:52 | |
| fungi | SergeyLukjanov: yes, when it tries to merge it will at that point see the -2 and refuse to do so | 14:52 |
| stevemar | dstanek, the first part of your comment on my refactor (soon to be fix) for base_url in trusts had me surprised | 14:52 |
| SergeyLukjanov | fungi, awesome | 14:52 |
| *** cnesa has joined #openstack-dev | 14:52 | |
| *** dvarga has joined #openstack-dev | 14:52 | |
| *** achampion has joined #openstack-dev | 14:53 | |
| dims | saschpe_, thanks a ton | 14:53 |
| fungi | SergeyLukjanov: on the to do list for zuul is to have it notice -2 cdrv or 0 aprv events in the stream for enqueued changes and abort their jobs immediately instead of waiting until merge time to check | 14:53 |
| fungi | abort their jobs and completely dequeue them | 14:53 |
| SergeyLukjanov | fungi, yup, good idea | 14:54 |
| *** pschaef has joined #openstack-dev | 14:54 | |
| *** rcleere has quit IRC | 14:54 | |
| *** yamahata has quit IRC | 14:55 | |
| *** zaitcev has joined #openstack-dev | 14:57 | |
| YorikSar | SergeyLukjanov: In fact, now you can bring back you +2 and +A since we're agreed that we're ready to land that change. | 14:57 |
| *** jamieh has quit IRC | 14:57 | |
| *** dims has quit IRC | 14:57 | |
| stevemar | dstanek, i'm running some of the trust tests in v3 auth on master branch now, and seeing the correct behaviour for trusts self link... i'm wondering if you did anything different? | 14:58 |
| *** YorikSar_ has joined #openstack-dev | 14:59 | |
| SergeyLukjanov | YorikSar, I think that we've agreed on merging timeout and known_hosts | 14:59 |
| SergeyLukjanov | YorikSar, reading scroll back, too many chats in parallel | 14:59 |
| *** xga has joined #openstack-dev | 14:59 | |
| *** browne has joined #openstack-dev | 15:00 | |
| *** jamieh has joined #openstack-dev | 15:00 | |
| *** andreaf has quit IRC | 15:00 | |
| *** doug-fish has joined #openstack-dev | 15:00 | |
| *** dims has joined #openstack-dev | 15:00 | |
| *** YorikSar has quit IRC | 15:00 | |
| *** YorikSar_ is now known as YorikSar | 15:01 | |
| *** vijendar has joined #openstack-dev | 15:01 | |
| *** otherwiseguy has joined #openstack-dev | 15:01 | |
| *** yamahata has joined #openstack-dev | 15:02 | |
| YorikSar | SergeyLukjanov: Ok, gtg now. I'll take a look at changes again from home. | 15:02 |
| sdague | so is it common for keystone reviews to sit for 2 weeks with a +2 on them without any other core votes? | 15:02 |
| dolphm | sdague: link? | 15:02 |
| sdague | https://review.openstack.org/#/c/69218/ and https://review.openstack.org/#/c/69219/ | 15:03 |
| *** alex_xu has quit IRC | 15:03 | |
| *** byeager has joined #openstack-dev | 15:03 | |
| *** tsekiyama has joined #openstack-dev | 15:04 | |
| *** danielbruno has joined #openstack-dev | 15:04 | |
| *** jnoller has joined #openstack-dev | 15:05 | |
| *** mohits has joined #openstack-dev | 15:06 | |
| *** mgagne has joined #openstack-dev | 15:06 | |
| *** hartsocks has quit IRC | 15:07 | |
| *** edmund has joined #openstack-dev | 15:07 | |
| stevemar | dstanek, i see the bug now, it's only in list trusts | 15:08 |
| *** kenperkins has joined #openstack-dev | 15:09 | |
| *** arosen has quit IRC | 15:09 | |
| dolphm | gyee: you +2'd this right before a rebase https://review.openstack.org/#/c/68548/ | 15:09 |
| *** kolesovdv has quit IRC | 15:10 | |
| *** dvarga is now known as dvarga|away | 15:10 | |
| *** dvarga|away is now known as dvarga | 15:10 | |
| dstanek | hi stevemar | 15:11 |
| dstanek | stevemar: that was a wierd one | 15:11 |
| *** mdenny has joined #openstack-dev | 15:11 | |
| stevemar | dstanek, yeah, interesting | 15:11 |
| stevemar | dstanek, i don't think any of the extensions do the links properly | 15:12 |
| stevemar | dstanek, more work then | 15:12 |
| *** bhuvan has joined #openstack-dev | 15:13 | |
| *** Tross has joined #openstack-dev | 15:13 | |
| *** eglynn is now known as eglynn-afk | 15:13 | |
| *** max_lobur is now known as max_lobur_afk | 15:14 | |
| *** CaptTofu has quit IRC | 15:17 | |
| *** Gordonz has joined #openstack-dev | 15:18 | |
| *** Gordonz has quit IRC | 15:18 | |
| *** Gordonz has joined #openstack-dev | 15:18 | |
| *** sarob has joined #openstack-dev | 15:19 | |
| *** devoid has joined #openstack-dev | 15:19 | |
| stevemar | bknudson, can you review https://review.openstack.org/#/c/72544/ it should get rid of the last errors we were seeing in doc building | 15:20 |
| *** thouveng has quit IRC | 15:22 | |
| *** MaxV has quit IRC | 15:23 | |
| *** MaxV has joined #openstack-dev | 15:23 | |
| *** kbrierly has joined #openstack-dev | 15:23 | |
| *** sarob has quit IRC | 15:24 | |
| *** rcleere has joined #openstack-dev | 15:24 | |
| *** dbalog has joined #openstack-dev | 15:24 | |
| ayoung | jdennis, I'm working on compressing PKI tokens. In doing so, I want to redo the test for a PKI token. I want to replace the check for "MII" at the front with a base64 decdoe | 15:25 |
| ayoung | decode | 15:25 |
| *** FunnyLookinHat has quit IRC | 15:25 | |
| ayoung | jdennis, as you know, we do a non-standard transform to go from base64 standard encoding to an URL safe format (replace / with -) | 15:26 |
| *** jpomero has joined #openstack-dev | 15:26 | |
| BLZbubba | how useful is the Baremetal system in Havana? | 15:26 |
| ayoung | I'm trying to determine if I can safely ignore that when testing the token: both characters show up as valid Base64 if I use pythons base64.urlsafe_b64decode, just in once case the decoded doc would be garbage. | 15:27 |
| *** MaxV has quit IRC | 15:28 | |
| ayoung | However, I'm a little reluctant to just blindly assume - should be / and transform the document. | 15:28 |
| BLZbubba | and also, is there an easy way to snapshot and redeploy a whole tenant, and give each tenant its own dns subdomain? | 15:28 |
| *** Shaan7 has joined #openstack-dev | 15:28 | |
| *** Shaan7 has joined #openstack-dev | 15:28 | |
| ayoung | jdennis, If, instead of the openssl cms producing PEM, I could produce DER and encode using base64.urlsafe_b64encode. If I do that, I assume I will never get a '/' | 15:29 |
| *** tzabal has joined #openstack-dev | 15:29 | |
| *** Tross has quit IRC | 15:29 | |
| ayoung | I'm guessing that what I need to do is to decode assuming - means - (urlsafe) then test if the underlying doc is ASN 1. If it is not ASN1, do the transform from - to / and decode standard. | 15:30 |
| ayoung | dstanek, stevemar ^^ I was asking jdennis since he has some familiarity here, but if either of you guys know, please chime in | 15:31 |
| *** carl_baldwin has joined #openstack-dev | 15:31 | |
| jdennis | ayoung: maybe I'm not following you fully, sounds like you're trying to invent a heuristic that is not robust | 15:32 |
| *** MaxV has joined #openstack-dev | 15:32 | |
| *** zoresvit has joined #openstack-dev | 15:33 | |
| *** sballe has joined #openstack-dev | 15:33 | |
| jdennis | ayoung: the absence of a single alphabet character does not tell you anything about the alphabet in use | 15:33 |
| jdennis | ayoung: also, not sure what you mean by PEM vs. DER, I don't see how that changes anything | 15:34 |
| ayoung | jdennis, OK, let me start with the PEM vs DER thing | 15:35 |
| jdennis | ayoung: maybe you could start with what you're trying to accomplish | 15:35 |
| *** colinmcnamara has joined #openstack-dev | 15:35 | |
| ayoung | Current code to create a token signed the docuemnt in PEM format, then strips the header and foorter, and converts the / to the - character | 15:35 |
| ayoung | if I want to test to see if this is a valid token, I don't need to convert all the way back to PEM format (add the headers) I just need to: | 15:36 |
| *** sergmelikyan has quit IRC | 15:36 | |
| ayoung | convert - back to / and base64 decode | 15:36 |
| *** samuelbercovici has quit IRC | 15:36 | |
| *** tkammer has quit IRC | 15:37 | |
| *** dkehn__ has joined #openstack-dev | 15:37 | |
| *** coolsvap has joined #openstack-dev | 15:37 | |
| jdennis | ayoung: what is your definition of a valid token? | 15:38 |
| *** dkehn_ has quit IRC | 15:38 | |
| ayoung | jdennis, ah, good point, so the check is_asn1_token is used to distinguish between UUID token and a PKI based token | 15:38 |
| *** atiwari has joined #openstack-dev | 15:39 | |
| ayoung | one must be validated via online lookup, the other can be checked in process | 15:39 |
| *** nelsnelson has joined #openstack-dev | 15:39 | |
| *** chandan_kumar has quit IRC | 15:39 | |
| *** chandan_kumar has joined #openstack-dev | 15:40 | |
| ayoung | so I am trying to make changes in the client code to check the validity that will work on top of the patch you saw to replace the server keystone.common.cms with the client keystoneclient.common.cms | 15:40 |
| jdennis | ayoung: ok, so are you asking if is_asn1_token implemented correctly or if once you know it's a PKI based token how to validate it? | 15:40 |
| ayoung | jdennis this is a case of stepping into messy code and figuring out the right way to fix it. I had a big patch that was origianlly for compressing the PKI tokens, but if I make them smaller, the is_asn1_token check for "MII" fails | 15:42 |
| ayoung | soi I need to clean up that function | 15:42 |
| *** JoshuaG_AIM has joined #openstack-dev | 15:42 | |
| ayoung | MII is the length of the document, but with compression, the tokens will be shorter, and might not start with MII anymore | 15:42 |
| *** Tross has joined #openstack-dev | 15:42 | |
| ayoung | so...first thing is to clean up the check is_asn1_token. I think the robust-but-expensive approach would be: | 15:43 |
| *** dkehn_ has joined #openstack-dev | 15:43 | |
| ayoung | 1. Do a base64 decode followed by an ASN1 parse. If that succeeds, it is an asn1 token | 15:44 |
| jdennis | ayoung: well right :-) Isn't this what I said earlier? Aren't you trying to invent a hueristic, one that's difficult to implement robustly? Why not tag the data and eliminate fancy guessing? | 15:44 |
| JoshuaG_AIM | Does anyone know how to attach a software vlan to a hardware vlan in Open Stack? | 15:44 |
| ayoung | jdennis, because I need to deal with existing tokens as well. We need to be able to transition | 15:44 |
| bknudson | ayoung: existing tokens have MII | 15:44 |
| bknudson | MII should still work | 15:44 |
| bknudson | "work" in quotes | 15:45 |
| bknudson | unless a UUID token happens to start with MII | 15:45 |
| ayoung | bknudson, not possible | 15:45 |
| ayoung | UUID is hexadecimal only | 15:45 |
| ayoung | uuid.uuid4().hex | 15:45 |
| ayoung | bknudson, there is also an outstanding bug for tokens that are too large they do not start with MII, but MIJ | 15:46 |
| bknudson | ayoung: that should make picking a prefix easy... pick something outside of the hex digit range | 15:46 |
| *** rgerganov_ has joined #openstack-dev | 15:46 | |
| ayoung | bknudson, I'm trying to avoid that | 15:46 |
| *** dkehn__ has quit IRC | 15:47 | |
| *** dkehn__ has joined #openstack-dev | 15:47 | |
| *** tsekiyam_ has joined #openstack-dev | 15:47 | |
| jdennis | ayoung: I think your basic approach is right, try decompressing, then apply the test. But really this is why unknown, unstructured data is always a bad idea | 15:47 |
| *** rtheis_ has joined #openstack-dev | 15:47 | |
| bknudson | ayoung: what's wrong with defining a prefix? | 15:47 |
| *** cagrev_ has joined #openstack-dev | 15:47 | |
| *** aswadrangnekar1 has joined #openstack-dev | 15:47 | |
| *** mflobo_ has joined #openstack-dev | 15:47 | |
| ayoung | bknudson, it takes something that is in a standard file format (base64 encode) and morphs it into a hybrid: part Base64, part text | 15:47 |
| *** sweston has joined #openstack-dev | 15:48 | |
| *** CaptTofu has joined #openstack-dev | 15:48 | |
| bknudson | ayoung: use a different header | 15:48 |
| ayoung | it means that you can't just take a token and run it through a CLI to get the underlying data...which jdennis found out the hard way | 15:48 |
| *** iartaris` has joined #openstack-dev | 15:48 | |
| ayoung | bknudson, so..I'll do a header if necessary | 15:48 |
| *** MIDENN_ has joined #openstack-dev | 15:48 | |
| *** MIDENN_ has quit IRC | 15:48 | |
| *** sarob has joined #openstack-dev | 15:48 | |
| *** dstanek_afk has joined #openstack-dev | 15:48 | |
| *** prad_ has quit IRC | 15:49 | |
| *** MIDENN_ has joined #openstack-dev | 15:49 | |
| bknudson | the token will be like G-{base64-encoded} | 15:49 |
| *** xarses has quit IRC | 15:49 | |
| ayoung | part of this is that I am trying to break the large patch from last night into smaller, reviewable chunks that will also individually pass the gate | 15:49 |
| jdennis | ayoung: I think it's way better to use a header or prefix or whatever you want to call it, common usage is {xxx} fwiw | 15:49 |
| bknudson | so you run the part after G- through CLI | 15:49 |
| *** davidlenwell_ has joined #openstack-dev | 15:49 | |
| *** gcha has quit IRC | 15:49 | |
| ayoung | But that doesn't work for existing tokens, and dolphm objected to defining two different file formate | 15:50 |
| ayoung | formas | 15:50 |
| ayoung | ugh | 15:50 |
| ayoung | anyway | 15:50 |
| *** cfriesen_ has joined #openstack-dev | 15:50 | |
| *** MaxV_ has joined #openstack-dev | 15:50 | |
| *** galstrom is now known as galstrom_zzz | 15:50 | |
| JoshuaG_AIM | Does anyone know how to attach a software vlan to a hardware vlan in Open Stack? | 15:51 |
| *** gcha has joined #openstack-dev | 15:51 | |
| *** gcha has quit IRC | 15:51 | |
| *** carlp has joined #openstack-dev | 15:51 | |
| *** jprovazn has quit IRC | 15:51 | |
| *** cdub_ has joined #openstack-dev | 15:51 | |
| *** SpamapS_ has joined #openstack-dev | 15:51 | |
| ayoung | The mistake I made was doing my own base64 url-safe approach as opposed to using a standard one. | 15:51 |
| *** jmontemayor has joined #openstack-dev | 15:52 | |
| *** yeylon__ has quit IRC | 15:52 | |
| *** gilliard_ has joined #openstack-dev | 15:52 | |
| bknudson | why do PKI tokens need to be url safe? | 15:52 |
| *** jp_at_hp1 has joined #openstack-dev | 15:52 | |
| jdennis | ayoung: we really should get the base64utils and pemutils patch in | 15:52 |
| ayoung | bknudson, because they are sent over the wire | 15:52 |
| bknudson | ayoung: in a url? | 15:53 |
| *** atiwari_ has joined #openstack-dev | 15:53 | |
| ayoung | jdennis, are they submitted against python-keystoneclient | 15:53 |
| *** gcha has joined #openstack-dev | 15:53 | |
| ayoung | bknudson, in a header | 15:53 |
| *** sarob has quit IRC | 15:53 | |
| *** yaguang has quit IRC | 15:53 | |
| *** kgriffs_afk is now known as kgriffs | 15:53 | |
| ayoung | bknudson, curl etc barfs if they are not url safe | 15:53 |
| bknudson | seems like they need to be header-safe and not url safe? | 15:53 |
| ayoung | they get misparsed | 15:53 |
| ayoung | same diff | 15:54 |
| ayoung | bknudson, the utility is called URL safe | 15:54 |
| *** jamespag` has joined #openstack-dev | 15:54 | |
| *** jamespag` has joined #openstack-dev | 15:54 | |
| ayoung | base64.urlsafe_b64decode will produce different output than base64.b64decode( | 15:54 |
| *** AlanClark has joined #openstack-dev | 15:54 | |
| ayoung | >>> print raw2 | 15:55 |
| ayoung | ---------------------------------------- | 15:55 |
| ayoung | so that would get | 15:55 |
| ayoung | >>> base64.urlsafe_b64decode(raw2) | 15:55 |
| ayoung | '\xfb\xef\xbe\xfb\xef\xbe\xfb\xef\xbe\xfb\xef\xbe\xfb\xef\xbe\xfb\xef\xbe\xfb\xef\xbe\xfb\xef\xbe\xfb\xef\xbe\xfb\xef\xbe' | 15:55 |
| ayoung | but | 15:55 |
| ayoung | >>> base64.b64decode(raw2) | 15:55 |
| ayoung | '' | 15:55 |
| jdennis | ayoung: pemutils and base64utils was submitted to keystone, but now it sounds like were moving the common code to keystone-client | 15:55 |
| *** thedodd has joined #openstack-dev | 15:55 | |
| *** rgerganov_ has quit IRC | 15:55 | |
| jdennis | bknudson: there is a whole write-up on this issue in my patches | 15:56 |
| *** prad_ has joined #openstack-dev | 15:56 | |
| ayoung | both are considered valid base64. Just the underlying asn1 would be corrupted if the wrong one were used. I think that for "is_asn1_token" I can ignore the difference | 15:56 |
| bknudson | ayoung: ok. I don't have any objection to use base64.urlsafe_b64encode ... doesn't seem necessary for a header, though. | 15:56 |
| *** dkehn_ has quit IRC | 15:56 | |
| *** atiwari has quit IRC | 15:56 | |
| *** MaxV has quit IRC | 15:56 | |
| *** mdenny has quit IRC | 15:56 | |
| *** tsekiyama has quit IRC | 15:56 | |
| *** mkollaro has quit IRC | 15:56 | |
| *** cfriesen has quit IRC | 15:56 | |
| *** mflobo has quit IRC | 15:56 | |
| *** dkranz has quit IRC | 15:56 | |
| *** cagrev has quit IRC | 15:56 | |
| *** rtheis has quit IRC | 15:56 | |
| *** dstanek has quit IRC | 15:56 | |
| *** jay-lau-513 has quit IRC | 15:56 | |
| *** tmclaugh[work] has quit IRC | 15:56 | |
| *** markmc has quit IRC | 15:56 | |
| *** jruzicka has quit IRC | 15:56 | |
| *** martyntaylor has quit IRC | 15:56 | |
| *** qba73 has quit IRC | 15:56 | |
| *** jp_at_hp has quit IRC | 15:56 | |
| *** ygbo has quit IRC | 15:56 | |
| *** bvandenh has quit IRC | 15:56 | |
| *** iartarisi has quit IRC | 15:56 | |
| *** I159 has quit IRC | 15:56 | |
| *** viktors has quit IRC | 15:56 | |
| *** bauzas has quit IRC | 15:56 | |
| *** aswadrangnekar has quit IRC | 15:56 | |
| *** jaypipes has quit IRC | 15:56 | |
| *** alop has quit IRC | 15:56 | |
| *** erkules has quit IRC | 15:56 | |
| *** cdub has quit IRC | 15:56 | |
| *** coasterz has quit IRC | 15:56 | |
| *** nplanel has quit IRC | 15:56 | |
| *** d0ugal has quit IRC | 15:56 | |
| *** jamespage has quit IRC | 15:56 | |
| *** haleyb has quit IRC | 15:56 | |
| *** rgerganov has quit IRC | 15:56 | |
| *** SpamapS has quit IRC | 15:56 | |
| *** ogelbukh has quit IRC | 15:56 | |
| *** davidlenwell has quit IRC | 15:56 | |
| *** mordred has quit IRC | 15:56 | |
| *** gmoro has quit IRC | 15:56 | |
| *** gilliard has quit IRC | 15:56 | |
| ayoung | jdennis, can you resubmit them against client | 15:56 |
| ayoung | jdennis, can you resubmit them against client | 15:56 |
| *** zul has joined #openstack-dev | 15:56 | |
| *** devoid has quit IRC | 15:56 | |
| *** jay-lau-513 has joined #openstack-dev | 15:57 | |
| *** markwash has joined #openstack-dev | 15:57 | |
| ayoung | jdennis, BTW, thanks for the review on the "dedupe of CMS" I'll push that along | 15:57 |
| *** alop has joined #openstack-dev | 15:57 | |
| *** ifarkas has quit IRC | 15:58 | |
| *** pradeep1 has quit IRC | 15:58 | |
| *** d0ugal has joined #openstack-dev | 15:58 | |
| jdennis | bknudson: part of the problem we're trying to correct is that we invented another non-standard base64 alphabet when there was already an RFC defined alphabet with support already present in core python | 15:58 |
| ayoung | jdennis, ++ | 15:58 |
| *** yjiang51 has joined #openstack-dev | 15:59 | |
| ayoung | I'm trying to provide a transition strategy | 15:59 |
| *** chandan_kumar has quit IRC | 15:59 | |
| *** kmartin has joined #openstack-dev | 15:59 | |
| *** chandankumar_ has joined #openstack-dev | 15:59 | |
| *** irenab has joined #openstack-dev | 15:59 | |
| *** _rdas_ has joined #openstack-dev | 15:59 | |
| *** dims has quit IRC | 15:59 | |
| ayoung | jdennis, so to use the python library version, I want to do openssl cms -inform=DER and then base64.urlsafe_b64encode instead | 15:59 |
| *** nermina has quit IRC | 15:59 | |
| *** atiwari_ has quit IRC | 16:00 | |
| *** rdas has quit IRC | 16:00 | |
| ayoung | Of course, that is a server change, ATM | 16:00 |
| *** atiwari has joined #openstack-dev | 16:00 | |
| *** carlp has quit IRC | 16:00 | |
| *** markwash_ has joined #openstack-dev | 16:00 | |
| *** tmclaugh[work] has joined #openstack-dev | 16:01 | |
| *** cnesa has quit IRC | 16:01 | |
| *** Ruetobas has quit IRC | 16:01 | |
| *** coolsvap1 has joined #openstack-dev | 16:01 | |
| jdennis | ayoung: yes you want to use the urlsafe_b64{encode,decode} variants | 16:01 |
| *** markwash has quit IRC | 16:01 | |
| *** markwash_ is now known as markwash | 16:01 | |
| *** amotoki_ has joined #openstack-dev | 16:02 | |
| *** jckasper has joined #openstack-dev | 16:02 | |
| *** jaypipes has joined #openstack-dev | 16:02 | |
| *** gmoro has joined #openstack-dev | 16:02 | |
| *** ogelbukh has joined #openstack-dev | 16:02 | |
| *** ygbo has joined #openstack-dev | 16:02 | |
| ayoung | jdennis, so I'm trying to introduce as small a change at a time to do that. First is to make the client accept either format | 16:02 |
| *** markmc has joined #openstack-dev | 16:02 | |
| *** bvandenh has joined #openstack-dev | 16:02 | |
| *** qba73 has joined #openstack-dev | 16:02 | |
| *** jruzicka has joined #openstack-dev | 16:02 | |
| *** paragan has quit IRC | 16:03 | |
| *** erkules has joined #openstack-dev | 16:03 | |
| *** sarob has joined #openstack-dev | 16:03 | |
| *** dkranz has joined #openstack-dev | 16:03 | |
| *** haleyb has joined #openstack-dev | 16:03 | |
| *** Ruetobas has joined #openstack-dev | 16:03 | |
| *** coolsvap has quit IRC | 16:03 | |
| *** nplanel has joined #openstack-dev | 16:03 | |
| *** coasterz has joined #openstack-dev | 16:03 | |
| sdake | apparently something I was doing to get tox and testr rolling broke my env, and now I get http://paste.fedoraproject.org/76582/13922209 when installing a python program via setup | 16:03 |
| sdake | any tips? | 16:03 |
| *** martyntaylor has joined #openstack-dev | 16:04 | |
| *** viktors has joined #openstack-dev | 16:04 | |
| sdake | google has no answers | 16:04 |
| *** comay has quit IRC | 16:04 | |
| *** mordred has joined #openstack-dev | 16:04 | |
| *** I159 has joined #openstack-dev | 16:04 | |
| *** dstanek_afk is now known as dstanek | 16:04 | |
| *** iartaris` is now known as iartarisi | 16:04 | |
| jdennis | ayoung: I could sit down and try to figure out a robust way to determine which alphabet is in use, but I'm in the middle of the utf-8 fixes atm, | 16:05 |
| ayoung | jdennis, NP | 16:05 |
| *** bauzas has joined #openstack-dev | 16:05 | |
| ayoung | jdennis, just writing it out has made the issues clearer for me | 16:05 |
| jdennis | ayoung: why not solve the entire problem with a prefix, it makes everything much cleaner and is extensible | 16:05 |
| jdennis | ayoung: if you don't see a prefix then you know you're dealing with legacy and you only have 2 options | 16:06 |
| *** FunnyLookinHat has joined #openstack-dev | 16:06 | |
| jdennis | ayoung: if you see a prefix then the world is beautiful because you can key off the prefix and all problems go away | 16:06 |
| roaet | Anyone run into TypeError: dist must be a Distribution instance when installing things? (neutron in my case) | 16:07 |
| ayoung | jdennis, except that you now have a custom file format that none of the tools work on without know-how | 16:07 |
| jdennis | ayoung: what file format? we're not discussing a file format at all | 16:08 |
| *** sarob has quit IRC | 16:08 | |
| alex-foo | the git-review tests are completely unstable, i've tried 4 times to get them through now but at least 1 of 2.6, 2.7 or 3.3 will fail sporadically and it takes FORTY MINUTES to time out :/ | 16:08 |
| ayoung | jdennis, and are the a standard set of {prefxies} or do we just make something up? | 16:08 |
| *** Ruetobas has quit IRC | 16:08 | |
| jdennis | ayoung: what I'm talking about is a format for token id's, that's exclusive to openstack and will only ever been seen in protocol exchanges | 16:09 |
| ayoung | jdennis, I'm not certain about that. I'd like to avoid inventing my own file format if necesary, but I can see your point. | 16:10 |
| *** gcha has quit IRC | 16:10 | |
| jdennis | ayoung: the problem is we have a data item, the token_id, and it's a string that has to be interpreted, but there is no information on how to interpret the string | 16:10 |
| jdennis | what file format? | 16:10 |
| *** mrda_away is now known as mrda | 16:11 | |
| *** Ruetobas has joined #openstack-dev | 16:11 | |
| jdennis | where is the token_id ever stored in a file that is parsed later? | 16:11 |
| *** anniec has joined #openstack-dev | 16:11 | |
| *** gyee has quit IRC | 16:11 | |
| *** branen has joined #openstack-dev | 16:12 | |
| *** nosnos has quit IRC | 16:12 | |
| cyeoh | mikal: would you mind having a look at https://review.openstack.org/#/c/63566/ ? Has one +2 already | 16:13 |
| *** otherwiseguy has quit IRC | 16:13 | |
| *** nshaikh has left #openstack-dev | 16:14 | |
| *** xmltok has joined #openstack-dev | 16:14 | |
| jdennis | ayoung: if you're looking for an analogy then consider password hashes in ldap attributes, folks wold store raw hash values in the attribute | 16:14 |
| *** dims has joined #openstack-dev | 16:14 | |
| *** markmcclain has joined #openstack-dev | 16:14 | |
| *** kushal has quit IRC | 16:14 | |
| jdennis | ayoung: but that only worked if you knew apriori which hash was used and it prevented you from ever using other hashes | 16:15 |
| *** david-lyle has joined #openstack-dev | 16:15 | |
| *** rodrigods has quit IRC | 16:15 | |
| *** anniec_ has joined #openstack-dev | 16:16 | |
| *** anniec has quit IRC | 16:16 | |
| *** anniec_ is now known as anniec | 16:16 | |
| *** neeti has joined #openstack-dev | 16:16 | |
| *** sushil_ has quit IRC | 16:17 | |
| *** dhellmann_ is now known as dhellmann | 16:17 | |
| jdennis | ayoung: the solution adopted was to prefix each hash in a password attribute with {xxx} where xxx identifies the hash algorithm, if the hash starts with {*} then you strip off the prefix and use the remainder, and the prefix tells you exactly how to interpret the remainder, if there is no prefix then you're legacy and have to fall back to some default | 16:17 |
| *** AlexF has joined #openstack-dev | 16:18 | |
| *** mriedem has joined #openstack-dev | 16:18 | |
| ayoung | jdennis, then the comparable solution here would be to put the form of base64 encoding used in {} | 16:18 |
| jdennis | ayoung: that's the basic idea, but it's more than just base64, all base64 tells you is you've got binary data, not what kind of binary data, right? | 16:19 |
| *** tjones has joined #openstack-dev | 16:19 | |
| *** gyee has joined #openstack-dev | 16:19 | |
| jdennis | ayoung: there should be a prefix for UUID format as well | 16:19 |
| ayoung | jdennis, so...I end up putting all of the operations into the prefix | 16:19 |
| *** hartsocks has joined #openstack-dev | 16:20 | |
| ayoung | so what do I do? go all cobal like? | 16:20 |
| jdennis | ayoung: it's really simple, you're presented with a blob of data, you need something to tell you how to interpret that blob, or you have to invent rickity hueristics | 16:20 |
| *** pablosan has joined #openstack-dev | 16:21 | |
| ayoung | jdennis, {md5} or {sha1} are standard formats | 16:21 |
| *** pmathews has joined #openstack-dev | 16:21 | |
| *** zoresvit has quit IRC | 16:21 | |
| ayoung | the standard format I want here doesn't quite exist | 16:21 |
| ayoung | since openssl selected standard base64 instead of URL safe... | 16:21 |
| jdennis | ayoung: those are standard for the case of password hashes, but we're in our own domain here, we can define whatever we want | 16:22 |
| ayoung | or CMS rfs specified | 16:22 |
| ayoung | except that all I want is the DER file underneath | 16:22 |
| ayoung | and base64 is the way to make it so I can transprot it across the wire. | 16:22 |
| *** hartsocks has quit IRC | 16:22 | |
| ayoung | transport | 16:22 |
| devananda | jog0: hacking question - is there a general opinion in interpolating LOG messages vs passing in parameters? | 16:22 |
| jdennis | ayoung: I think you're too deep in the details, pop up a level | 16:23 |
| *** alex_xu has joined #openstack-dev | 16:24 | |
| *** thomasem_ is now known as thomasem | 16:24 | |
| *** Mandell has quit IRC | 16:24 | |
| *** tanisdl has joined #openstack-dev | 16:25 | |
| *** otherwiseguy has joined #openstack-dev | 16:25 | |
| *** beagles is now known as beagles_brb | 16:25 | |
| jdennis | ayoung: the token id is a blob of data, it has both formatting and content, you need to know it's formatting to interpret it's content | 16:25 |
| *** buzztroll has joined #openstack-dev | 16:26 | |
| *** tanisdl has joined #openstack-dev | 16:26 | |
| ayoung | jdennis, my goal is to get away from that. I want to be able to say "token is an urlsafe base64 encoded CMS document in DER format" | 16:26 |
| *** prad__ has joined #openstack-dev | 16:26 | |
| *** sushils has joined #openstack-dev | 16:27 | |
| jdennis | ayoung: you need something, either a prefix or some other attribute which is bound to it which tells you how to interpret the token id blob | 16:27 |
| ayoung | Then people can use base64 decode and CMS utilities. It makes the integration with other languages easy, and so forth. I don;t want to proviude multiple options, I want to atone for the sins of the past | 16:27 |
| *** thuc has joined #openstack-dev | 16:27 | |
| *** prad_ has quit IRC | 16:28 | |
| *** neelashah has joined #openstack-dev | 16:28 | |
| jog0 | devananda: passing in is prefered but not a big deal | 16:29 |
| jog0 | passing in is the standard logging way | 16:29 |
| jog0 | there is a reason why its considered better, but I forgot | 16:29 |
| jdennis | ayoung: but what happens when "token is an urlsafe base64 encoded CMS document in DER format" is no longer sufficient? you want to code for extensibility and robustness | 16:29 |
| ayoung | jdennis, at that point I will find a new profession. I'm thinking hot dog vendor | 16:30 |
| devananda | jog0: i'll take "not a big deal" to mean "not worth enforcing in hacking" :) | 16:30 |
| *** dvarga is now known as dvarga|away | 16:30 | |
| *** dvarga|away is now known as dvarga | 16:30 | |
| jdennis | ayoung: do you see the problem that migrating from UUID to PKI in the token id presented? it's the same problem and it's only going to get worse if not addressed | 16:31 |
| *** spzala has joined #openstack-dev | 16:31 | |
| ayoung | jdennis, yeah. I'll think about it. | 16:31 |
| *** sushils has quit IRC | 16:32 | |
| *** thuc has quit IRC | 16:32 | |
| *** sushils has joined #openstack-dev | 16:33 | |
| *** markmcclain has quit IRC | 16:33 | |
| *** markmcclain has joined #openstack-dev | 16:34 | |
| *** colinmcnamara has quit IRC | 16:34 | |
| ttx | dhellmann: my rootwrapectomy from oslo-incubator fails tests because I removed the rootwrap.conf in etc/ and tox runs generate_sample.sh over a non-existent directory now | 16:36 |
| *** MaxV_ has quit IRC | 16:36 | |
| ttx | dhellmann: should I just remove the call to generate_sample in tox.ini ? and/or remove tools/config/generate_sample.sh from oslo-incubator while I'm at it ? | 16:37 |
| ttx | (context: https://review.openstack.org/#/c/72968/1) | 16:37 |
| *** hartsocks has joined #openstack-dev | 16:37 | |
| *** sushils has quit IRC | 16:38 | |
| *** sushils has joined #openstack-dev | 16:38 | |
| ttx | dhellmann: I suspect it's useless now, but could use your confirmation :) | 16:39 |
| dhellmann | ttx: looking | 16:40 |
| *** MaxV has joined #openstack-dev | 16:40 | |
| *** byeager has quit IRC | 16:40 | |
| *** sweston has quit IRC | 16:40 | |
| dhellmann | ttx: if the only thing generate_sample.sh was doing was checking the sample config for rootwrap, that should be done in the new oslo.rootwrap repo | 16:41 |
| ttx | dhellmann: except we don't ship configurations there. They are shipped in the consuming projects (nova, cinder, neutron) | 16:41 |
| *** CaptTofu has quit IRC | 16:41 | |
| dhellmann | ttx: good point | 16:41 |
| *** xga has quit IRC | 16:42 | |
| *** xga has joined #openstack-dev | 16:42 | |
| ttx | dhellmann: just wondering if oslo-incubator/tools/config/* was resued as the sourcve for all the code copies around | 16:42 |
| *** nacim has quit IRC | 16:42 | |
| ttx | dhellmann: I suspect not | 16:42 |
| dhellmann | ttx: tools/config/generate_sample.sh probably needs to move to oslo.config | 16:42 |
| dhellmann | yeah | 16:42 |
| dhellmann | I think it is? | 16:42 |
| ttx | I can keep it in and just remove the call to it in tox | 16:43 |
| *** mikeoutland has joined #openstack-dev | 16:43 | |
| dhellmann | ttx: the difference between that file and the one in nova corresponds to changes that would be made by syncing (s/oslo/nova/) | 16:43 |
| dhellmann | ttx: what is etc/openstack.conf.sample? is that used by the tests? | 16:44 |
| *** coolsvap1 has quit IRC | 16:44 | |
| *** sandywalsh has joined #openstack-dev | 16:44 | |
| *** aditirav has quit IRC | 16:44 | |
| *** FunnyLookinHat has quit IRC | 16:44 | |
| dhellmann | oh, nevermind, that's the output file | 16:45 |
| *** amuller has quit IRC | 16:45 | |
| ttx | yep, not in repo | 16:45 |
| ttx | commit e013ac8980755b416f529c9d6b87ecfa61d31b1a says "Adds sample config file generation script, Having this script placed in oslo will let projects avoid implementing their own sample generation script." | 16:45 |
| ttx | dhellmann: I'll just remove the call to generate_sample from tox.ini and keep it in. | 16:46 |
| dhellmann | ttx: markmc wanted the generator to go through the incubator before moving to oslo.config | 16:46 |
| dhellmann | ttx: yes, I think that's the right approach | 16:46 |
| ttx | right | 16:46 |
| ttx | dhellmann: thanks for your advice! | 16:46 |
| dhellmann | I'm still trying to understand why that missing rootwrap file is causing a problem | 16:46 |
| ttx | it's actually now missing the etc/ directory altogether | 16:47 |
| ttx | since that was the only file in it | 16:47 |
| dhellmann | ah, right | 16:47 |
| *** colinmcnamara has joined #openstack-dev | 16:47 | |
| dhellmann | ok, so it can't *write* to that directory | 16:47 |
| YorikSar | fungi, SergeyLukjanov: Looks like one of them finally made it. Can you approve https://review.openstack.org/70685 again? | 16:47 |
| dhellmann | it's not trying to read the file, it's trying to write to the parent directory | 16:47 |
| dhellmann | ttx: ok, I'm satisfied, removing the line from tox.ini is the right fix | 16:48 |
| *** ayoung has quit IRC | 16:48 | |
| *** mflobo_ has quit IRC | 16:48 | |
| ttx | done | 16:49 |
| *** dtantsur has quit IRC | 16:50 | |
| *** abhirc has joined #openstack-dev | 16:50 | |
| *** _cjones_ has joined #openstack-dev | 16:51 | |
| *** bauzas has quit IRC | 16:53 | |
| *** xarses has joined #openstack-dev | 16:53 | |
| *** CaptTofu has joined #openstack-dev | 16:54 | |
| *** nacim has joined #openstack-dev | 16:54 | |
| *** FunnyLookinHat has joined #openstack-dev | 16:56 | |
| *** tmclaugh[work] has quit IRC | 16:57 | |
| YorikSar | fungi: Thank you. | 16:58 |
| *** pschaef has quit IRC | 16:59 | |
| *** jsavak has joined #openstack-dev | 16:59 | |
| *** coolsvap has joined #openstack-dev | 17:00 | |
| jordanP | Hi guys. Could a friendly tempest dev make a review on "Scenario : Start instances using fixed network when possible" : https://review.openstack.org/#/c/68904/ thanks :) | 17:00 |
| *** tqtran has joined #openstack-dev | 17:00 | |
| SergeyLukjanov | YorikSar, heh, it looks like it's quite difficult to push timeout patch :) | 17:01 |
| *** xga has quit IRC | 17:01 | |
| *** xga has joined #openstack-dev | 17:02 | |
| raildo | gyee: \o | 17:02 |
| SergeyLukjanov | jordanP, probably, you'd like to go to the #openstack-qa channel | 17:02 |
| *** networks_ has joined #openstack-dev | 17:02 | |
| *** mohits_ has joined #openstack-dev | 17:02 | |
| jordanP | SergeyLukjanov, ok thanks | 17:02 |
| *** joesavak has quit IRC | 17:02 | |
| gyee | raildo, here | 17:03 |
| *** devoid has joined #openstack-dev | 17:03 | |
| *** thomasem_ has joined #openstack-dev | 17:03 | |
| raildo | gyee: tellesnobrega and I have a doubt, in the solution of nested projects at Keystone and we would appreciate if you could help us. | 17:03 |
| *** avishay is now known as avishay_away | 17:04 | |
| *** thomasem_ has quit IRC | 17:04 | |
| raildo | We added in the table "project" a column "parent_project_id", that is a foreign key that references the table itself the column "id" of the parent project. | 17:04 |
| raildo | However the community decided that the project root "OpenStack" would not be a project in itself, is only one variable, so this will not work contraint. | 17:05 |
| gyee | raildo, that won't work | 17:05 |
| *** e0ne has quit IRC | 17:05 | |
| raildo | Do you have any suggestions for this problem? Remove the constraint, or some other way of upgrading the table. | 17:05 |
| *** thomasem has quit IRC | 17:05 | |
| gyee | foreign key can't be nullable I think | 17:06 |
| *** mohits has quit IRC | 17:06 | |
| *** nmagnezi has quit IRC | 17:06 | |
| tellesnobrega | gyee: they can | 17:06 |
| *** _rdas_ has quit IRC | 17:06 | |
| tellesnobrega | gyee: they can't be different from the values of the referenced column, but it can be null | 17:06 |
| gyee | tellesnobrega, so you can set parent_project_id to null to indicate root? | 17:07 |
| *** kenperkins has quit IRC | 17:07 | |
| *** kenperkins has joined #openstack-dev | 17:08 | |
| jog0 | devananda: yeah, although I don't remember the reason why its better | 17:08 |
| jog0 | if it was significantly better then maybe it is | 17:08 |
| *** dkehn__ is now known as dkehn_ | 17:08 | |
| *** Oneiroi has quit IRC | 17:08 | |
| *** henrynash has quit IRC | 17:09 | |
| raildo | I agree, but the idea of putting this foreign key, and the root project, for projects that do not have parents, this column will not be null. | 17:09 |
| *** jgallard has quit IRC | 17:09 | |
| *** evgenyf has quit IRC | 17:09 | |
| *** dvarga is now known as dvarga|away | 17:09 | |
| *** dvarga|away is now known as dvarga | 17:09 | |
| *** nagyz has quit IRC | 17:09 | |
| tellesnobrega | gyee: hum. thinking here | 17:09 |
| tellesnobrega | gyee: the structure we are using root -> domains -> projectA -> projectB | 17:10 |
| *** sushils has quit IRC | 17:10 | |
| *** marcoemorais has joined #openstack-dev | 17:10 | |
| tellesnobrega | so projectA should have Null to indicate that its the first of the hierarchy? | 17:10 |
| *** markmc has quit IRC | 17:11 | |
| tellesnobrega | or it should have the root project id? | 17:11 |
| tellesnobrega | it can't have the domain id | 17:11 |
| *** neeti has quit IRC | 17:11 | |
| *** kenperkins_ has joined #openstack-dev | 17:11 | |
| *** sushils has joined #openstack-dev | 17:11 | |
| devananda | jog0: performance of string interpolation for LOG messages that are below log threshold? | 17:11 |
| *** abhirc has quit IRC | 17:12 | |
| tellesnobrega | all projects in the tree will be contained by a domain, and all of them reference the domain, and the first project of the hierarchy should have a parent_project_id pointing to the root project or we can have Null | 17:12 |
| jog0 | devananda: that sounds right | 17:13 |
| *** hdd_ has quit IRC | 17:14 | |
| *** SumitNaiksatam has quit IRC | 17:14 | |
| tellesnobrega | i think having the root id is more readable, but in code it doesnt really make any difference. The change would be. if i use root project id i would have to remove the FK constraint, if i use Null that wont be necessary | 17:14 |
| *** sushils has quit IRC | 17:14 | |
| *** xga has quit IRC | 17:15 | |
| *** beagles_brb is now known as beagles | 17:15 | |
| *** kenperkins has quit IRC | 17:15 | |
| *** xga has joined #openstack-dev | 17:15 | |
| gyee | tellesnobrega, raildo, not sure if implementing mix data structure in a tree is a good idea | 17:16 |
| *** mikeoutland has quit IRC | 17:16 | |
| *** Tross has quit IRC | 17:17 | |
| tellesnobrega | gyee: do you mean having projects and domains? | 17:17 |
| *** zul has quit IRC | 17:17 | |
| gyee | unless you want to define a generic data model and have a container type attribute | 17:17 |
| *** AlexF has quit IRC | 17:17 | |
| *** joesavak has joined #openstack-dev | 17:17 | |
| gyee | container type indicates whether it is a project or domain container | 17:18 |
| gyee | container would be implemented as a tree | 17:18 |
| *** zul has joined #openstack-dev | 17:18 | |
| *** zul has quit IRC | 17:18 | |
| *** zul has joined #openstack-dev | 17:18 | |
| gyee | tellesnobrega, root->container->container->container etc | 17:18 |
| tellesnobrega | i see | 17:18 |
| gyee | container = {type, id, etc} | 17:18 |
| *** galstrom_zzz is now known as galstrom | 17:19 | |
| dolphm | anyone know Mikhail Durnosvistov<mdurnosvistov@mirantis.com>'s IRC handle? | 17:19 |
| *** mst89 has joined #openstack-dev | 17:20 | |
| tellesnobrega | i see that this may be the best solution so far, but how much of a change would this cause in keystone? i guess this would have to be discussed for quite a while | 17:20 |
| *** jsavak has quit IRC | 17:20 | |
| *** anniec has quit IRC | 17:20 | |
| gyee | tellesnobrega, you mean like seismic changes? :) | 17:21 |
| dolphm | gyee: from an implementation perspective, "domain" is just a project where the parent_project_id is null; you don't have to complicate it beyond that | 17:21 |
| raildo | dolphm: I believe it is mdurnosvistov | 17:22 |
| gyee | dolphm, right, but implementing mix data structure in a tree is not fun | 17:22 |
| gyee | LDAP handle it well, not sure about SQL | 17:22 |
| annegentle | dolphm: ping | 17:22 |
| dolphm | raildo: ah, just not online right now then. thanks | 17:22 |
| dolphm | annegentle: o/ | 17:22 |
| annegentle | heya | 17:22 |
| *** praneshp has joined #openstack-dev | 17:22 | |
| raildo | dolphm: you're welcome | 17:22 |
| annegentle | so I'm hearing for deployers testing icehouse that v2 doesn't work nor does v3 api... and wondered what deprecated means for https://etherpad.openstack.org/p/icehouse-keystone-internal-apis? I must have missed the discussion | 17:23 |
| tellesnobrega | dolphm: we dont want to remove domains and use only projects | 17:23 |
| annegentle | for/from | 17:23 |
| dolphm | annegentle: that's alarming lol -- any links? | 17:23 |
| tellesnobrega | domains are important for us | 17:23 |
| annegentle | dolphm: I can find out more details but wanted to know the plan the plan :) | 17:23 |
| dolphm | annegentle: one second, that resulted in a bp | 17:23 |
| *** colinmcnamara has quit IRC | 17:24 | |
| *** gszasz has quit IRC | 17:24 | |
| *** I159 has quit IRC | 17:24 | |
| *** I159_ has quit IRC | 17:24 | |
| dolphm | annegentle: https://blueprints.launchpad.net/keystone/+spec/deprecated-as-of-icehouse | 17:24 |
| dolphm | annegentle: and to a lesser extent, https://blueprints.launchpad.net/keystone/+spec/list-limiting | 17:24 |
| annegentle | dolphm: Sam-I-Am is in the #openstack-doc channel and was asking about it, after testing icehouse Ubuntu packaging | 17:24 |
| *** praneshp has quit IRC | 17:25 | |
| *** xqueralt has quit IRC | 17:25 | |
| tellesnobrega | gyee: we only working on a prototype, the hierarchy itself would be only of projects and domains would contain a hierarchy, leaving the root out | 17:25 |
| *** godara has joined #openstack-dev | 17:25 | |
| annegentle | dolphm: are you getting input from users on the deprecation? I don't think it sounds like people are ready | 17:25 |
| annegentle | dolphm: also does deprecate mean it's available in icehouse but not available in juno? | 17:25 |
| annegentle | dolphm: I feel like I'm the canary in the coal mine, croaking with the last breath, that this is a BAD thing :) | 17:26 |
| dolphm | annegentle: everything deprecated in icehouse means it's guaranteed to be available in icehouse, but will alert consumers exactly when it's at risk to be removed (minimum +1 release, usually +2) | 17:26 |
| dolphm | annegentle: deprecations can be reversed, if it becomes clear later that it would be a mistake to drop support, etc | 17:26 |
| gyee | tellesnobrega, have fun, but you can't have a tree with no root :) | 17:26 |
| annegentle | dolphm: ok then the real comms will start after April or so, I still think you need a HUGE conversation with your install base because so far I'm not finding v3 | 17:27 |
| tellesnobrega | we will have a root | 17:27 |
| annegentle | dolphm: I would like some ML comms | 17:27 |
| *** godara_ has joined #openstack-dev | 17:27 | |
| tellesnobrega | just a second, i will show it graphically | 17:27 |
| annegentle | dolphm: at least (I know it was discussed at a summit and all, but I just don't think deployers know the ramifications) | 17:27 |
| dolphm | annegentle: i figured this would be a dedicated section in the release notes? | 17:27 |
| annegentle | dolphm: it's much much more invasive than that | 17:27 |
| dolphm | annegentle: discussion on *what's* being deprecated, or discussion on the process itself? | 17:28 |
| annegentle | dolphm: needs more investigation, upgrade notes | 17:28 |
| *** garyk has quit IRC | 17:28 | |
| annegentle | dolphm: what's being deprecated isn't clear for starters | 17:28 |
| *** ykhodork has joined #openstack-dev | 17:28 | |
| annegentle | dolphm: also what dependencies fall out | 17:28 |
| *** jdurgin has quit IRC | 17:28 | |
| annegentle | dolphm: like you don't get heat if you're running v2 | 17:28 |
| annegentle | dolphm: apparently | 17:28 |
| annegentle | that sort of thing. it's so invasive, install guide, user guides, ops guide | 17:28 |
| annegentle | just sending up the flag of concern :) | 17:29 |
| dolphm | annegentle: appreciated! | 17:29 |
| annegentle | now I have to drive to the office dolphm but thanks for "listening!" :) | 17:29 |
| *** godara has quit IRC | 17:29 | |
| *** godara_ is now known as godara | 17:29 | |
| dolphm | annegentle: ping me when you get there - i have a few follow up questions | 17:29 |
| annegentle | dolphm: I'll connect Sam-I-Am to ya (Matt Kassawar) for more details on his testing | 17:29 |
| *** bauzas has joined #openstack-dev | 17:29 | |
| *** carlp has joined #openstack-dev | 17:29 | |
| *** athomas has joined #openstack-dev | 17:29 | |
| *** jordanP has quit IRC | 17:30 | |
| *** carlp has quit IRC | 17:31 | |
| *** gokrokve has quit IRC | 17:32 | |
| *** kenperkins has joined #openstack-dev | 17:32 | |
| *** gokrokve has joined #openstack-dev | 17:32 | |
| *** questionmask_ has joined #openstack-dev | 17:34 | |
| *** Gordonz has quit IRC | 17:34 | |
| tellesnobrega | gyee: take a look at this http://oi61.tinypic.com/aaxksl.jpg | 17:35 |
| *** henrynash has joined #openstack-dev | 17:35 | |
| *** kenperkins_ has quit IRC | 17:36 | |
| *** bauzas has quit IRC | 17:36 | |
| gyee | tellesnobrega, that looks like an LDAP DIT | 17:36 |
| *** mmagr has quit IRC | 17:36 | |
| *** gokrokve has quit IRC | 17:37 | |
| *** lucasagomes is now known as lucas-afk | 17:37 | |
| *** bauzas has joined #openstack-dev | 17:38 | |
| *** danielbruno has quit IRC | 17:38 | |
| tellesnobrega | gyee: i guess so, but for now i think its a good solution. We dont need to eliminate domains, and we have the hierarchy of projects as proposed and the changes to keystone are not so seismic | 17:39 |
| *** qba73 has quit IRC | 17:39 | |
| *** tkay has joined #openstack-dev | 17:39 | |
| *** bdpayne has quit IRC | 17:39 | |
| *** rossella_s has quit IRC | 17:39 | |
| *** jdurgin has joined #openstack-dev | 17:40 | |
| *** beagles has quit IRC | 17:40 | |
| *** b3nt_pin has joined #openstack-dev | 17:41 | |
| *** b3nt_pin is now known as beagles | 17:41 | |
| *** ygbo has quit IRC | 17:42 | |
| stevemar | gordc, ping | 17:42 |
| *** Tross has joined #openstack-dev | 17:42 | |
| *** kris_h has joined #openstack-dev | 17:43 | |
| *** Tross has quit IRC | 17:43 | |
| *** Tross has joined #openstack-dev | 17:43 | |
| gordc | stevemar: got it | 17:44 |
| gyee | tellesnobrega, I don't wish to get into an academic debate right now, whatever good for your business is good for you :D | 17:44 |
| *** jtomasek has quit IRC | 17:45 | |
| tellesnobrega | gyee: im working with Tiago Martins and we really don't want to give up domains | 17:45 |
| gyee | tellesnobrega, understood | 17:47 |
| tellesnobrega | gyee: :D | 17:47 |
| *** gordc has quit IRC | 17:48 | |
| *** matsuhashi has quit IRC | 17:48 | |
| *** eharney has quit IRC | 17:49 | |
| *** dprince has quit IRC | 17:50 | |
| *** saju_m has quit IRC | 17:50 | |
| *** matsuhashi has joined #openstack-dev | 17:51 | |
| *** saju_m has joined #openstack-dev | 17:51 | |
| *** jsavak has joined #openstack-dev | 17:53 | |
| *** ala has quit IRC | 17:53 | |
| *** angdraug has joined #openstack-dev | 17:55 | |
| *** zz_ctlaugh is now known as ctlaugh | 17:55 | |
| *** matsuhashi has quit IRC | 17:55 | |
| *** mikeoutland has joined #openstack-dev | 17:56 | |
| *** comay has joined #openstack-dev | 17:56 | |
| *** dvarga has quit IRC | 17:56 | |
| *** bdpayne has joined #openstack-dev | 17:56 | |
| *** joesavak has quit IRC | 17:57 | |
| *** saju_m has quit IRC | 17:58 | |
| *** tsekiyam_ has quit IRC | 17:58 | |
| *** cburgess has quit IRC | 17:58 | |
| *** mohits has joined #openstack-dev | 17:59 | |
| *** danpb has quit IRC | 17:59 | |
| *** mandeep has joined #openstack-dev | 18:00 | |
| *** saju_m has joined #openstack-dev | 18:00 | |
| *** mohits has quit IRC | 18:00 | |
| *** mohits_ has quit IRC | 18:00 | |
| *** mohits has joined #openstack-dev | 18:01 | |
| *** AlexF has joined #openstack-dev | 18:01 | |
| *** jckasper has quit IRC | 18:02 | |
| *** pleia2 has quit IRC | 18:02 | |
| *** morganfainberg_Z has quit IRC | 18:02 | |
| shardy | dolphm: hi, I have a problem related to domains, devstack and the gate if you have a moment? | 18:03 |
| *** henrique has quit IRC | 18:03 | |
| *** chandankumar_ has quit IRC | 18:03 | |
| *** alexpilotti has joined #openstack-dev | 18:03 | |
| *** pleia2 has joined #openstack-dev | 18:03 | |
| *** rmk has quit IRC | 18:03 | |
| *** cburgess has joined #openstack-dev | 18:04 | |
| *** bvandenh has quit IRC | 18:05 | |
| *** ykhodork has quit IRC | 18:05 | |
| *** rmk has joined #openstack-dev | 18:06 | |
| *** morganfainberg_Z has joined #openstack-dev | 18:06 | |
| *** AlexF has quit IRC | 18:06 | |
| *** radez is now known as radez_g0n3 | 18:06 | |
| *** morganfainberg_Z is now known as morganfainberg | 18:06 | |
| *** morganfainberg has quit IRC | 18:07 | |
| *** morganfainberg has joined #openstack-dev | 18:07 | |
| *** bauzas has quit IRC | 18:07 | |
| *** eglynn-afk has quit IRC | 18:07 | |
| *** xga has quit IRC | 18:08 | |
| *** AlexF has joined #openstack-dev | 18:10 | |
| *** mandeep has quit IRC | 18:10 | |
| *** kgriffs is now known as kgriffs_afk | 18:12 | |
| *** mdurnosvistov_ has joined #openstack-dev | 18:13 | |
| *** tmclaugh[work] has joined #openstack-dev | 18:13 | |
| *** dperaza has joined #openstack-dev | 18:15 | |
| dolphm | shardy: o/ | 18:15 |
| shardy | dolphm: Hi | 18:16 |
| *** cadenzajon has joined #openstack-dev | 18:16 | |
| shardy | dolphm: I have an issue, heat needs to create a domain, and currently I have code in heat creating it as the heat stack user | 18:16 |
| shardy | but that works for local devstack, and not in the gate | 18:16 |
| *** pmathews has quit IRC | 18:16 | |
| shardy | I assume the gate tests are using the v3 policy sample? | 18:16 |
| *** MaxV has quit IRC | 18:16 | |
| dolphm | shardy: both policy files are tested | 18:17 |
| shardy | dolphm: Hmm, so can you think of some way we can create a domain in devstack, so I can set the ID in the heat config file? | 18:17 |
| dolphm | shardy: are you doing this in tempest? | 18:17 |
| shardy | dolphm: Yeah, it's a tempest test which is failing | 18:18 |
| dolphm | shardy: why does a domain need to be in heat's config file?? | 18:18 |
| dolphm | shardy: link? | 18:18 |
| shardy | dolphm: heat creates users in a heat-specific domain | 18:18 |
| shardy | dolphm: https://review.openstack.org/#/c/72763/ | 18:18 |
| shardy | dolphm: It's getting tripped up because I lookup a domain by name, and create it if it's missing | 18:19 |
| shardy | this was to avoid deployment pain, because there's no CLI support for domains yet | 18:19 |
| dolphm | shardy: so you're referring to "stack_user_domain" ? | 18:19 |
| dolphm | which i assume is intended to be a domain name? | 18:19 |
| shardy | dolphm: yes, exactly | 18:19 |
| *** gokrokve has joined #openstack-dev | 18:19 | |
| shardy | but we could move to setting and ID in the config | 18:20 |
| *** xgsa has quit IRC | 18:20 | |
| shardy | that would actually be preferable, but as I said, deployer pain | 18:20 |
| shardy | https://github.com/openstack/heat/blob/master/heat/common/heat_keystoneclient.py#L106 | 18:21 |
| dolphm | shardy: and it's a production expectation that he should be allowed to create domains? | 18:21 |
| dolphm | that heat* | 18:21 |
| shardy | dolphm: well ideally no, but it seemed a reasonable interim solution, given the lack of any admin tooling to create the domain at deployment time | 18:22 |
| *** chuck_ has joined #openstack-dev | 18:22 | |
| shardy | maybe we should just have a tools/create_heat_domain.py instead | 18:22 |
| dolphm | shardy: lack of tooling in devstack you mean? | 18:22 |
| *** mohits_ has joined #openstack-dev | 18:22 | |
| shardy | dolphm: lack of any CLI tool which can create domains | 18:22 |
| *** zul has quit IRC | 18:22 | |
| shardy | and in devstack, yes | 18:23 |
| dolphm | shardy: openstack domain-create | 18:23 |
| shardy | ooh, is that new? | 18:23 |
| dolphm | shardy: not at all | 18:23 |
| dolphm | shardy: last i checked it was broken against devstack's default catalog configuration | 18:23 |
| shardy | dolphm: Ok, maybe that's my mistake then, last time I looked python-openstackclient didn't handle domains | 18:24 |
| dolphm | shardy: --identity-api-version=3 | 18:24 |
| shardy | so if we can use that to create the domain in devstack and set the id in heat.conf, the problem is solved | 18:24 |
| shardy | aha! | 18:24 |
| shardy | that's what I was missing :) | 18:24 |
| shardy | thanks | 18:24 |
| dolphm | shardy: it needs to be the default :( | 18:24 |
| shardy | I was confused as it was calling v2 | 18:25 |
| *** Mandell has joined #openstack-dev | 18:25 | |
| dolphm | shardy: but again, i think openstackclient appears broken against devstack's default catalog; ayoung has a workaround that involves mutating catalog entries on the client side | 18:25 |
| *** iartarisi has quit IRC | 18:26 | |
| shardy | dolphm: Hmm, can we pass the endpoint explicitly via openstackclient? | 18:26 |
| *** mohits has quit IRC | 18:26 | |
| *** dprince has joined #openstack-dev | 18:26 | |
| dolphm | shardy: --os-endpoint, i believe | 18:27 |
| dtroyer | dolphm: I think the breakage is just that discovery doesn't work properly with the current SC default. ayoung's hack (or my incarnation of it) will let this work… | 18:27 |
| shardy | Maybe that's a viable immediate solution, so we can create the domain in devstack | 18:27 |
| *** colinmcnamara has joined #openstack-dev | 18:27 | |
| *** jpich has quit IRC | 18:27 | |
| *** safchain has quit IRC | 18:27 | |
| *** morganfainberg is now known as morganfainberg_Z | 18:27 | |
| *** sweston has joined #openstack-dev | 18:28 | |
| dtroyer | shardy: I think if you just set OS_IDENTITY_API_VERSION=3 (or use —os-identity-api-version) it should work... | 18:28 |
| *** harlowja_away is now known as harlowja | 18:28 | |
| shardy | dtroyer: Ok, thanks, I'll give that a try | 18:28 |
| *** dvarga has joined #openstack-dev | 18:28 | |
| shardy | dolphm: thanks for the info, sounds like a devstack patch will get things working for us | 18:29 |
| *** tsekiyama has joined #openstack-dev | 18:29 | |
| *** mohits has joined #openstack-dev | 18:30 | |
| *** otherwiseguy has quit IRC | 18:30 | |
| *** denis_ has joined #openstack-dev | 18:31 | |
| *** morganfainberg_Z is now known as morganfainberg | 18:32 | |
| morganfainberg | dolphm, ping | 18:32 |
| dolphm | morganfainberg: pong | 18:32 |
| *** dkehn__ has joined #openstack-dev | 18:32 | |
| morganfainberg | dolphm, are we still interested in moving KDS into it's own repo? | 18:32 |
| *** davidlenwell_ is now known as davidlenwell | 18:32 | |
| morganfainberg | dolphm, i think i have an infra changeset ready that would create the repo... think | 18:32 |
| morganfainberg | dolphm, but i wasn't sure what the direction on that was going to be | 18:33 |
| *** neelashah has quit IRC | 18:33 | |
| *** mohits_ has quit IRC | 18:33 | |
| *** nacim has quit IRC | 18:34 | |
| *** tsekiyama has quit IRC | 18:34 | |
| *** dkehn_ has quit IRC | 18:35 | |
| *** rossella_s has joined #openstack-dev | 18:35 | |
| mdurnosvistov_ | dolphm: ping :) | 18:36 |
| dolphm | mdurnosvistov_: o/ | 18:36 |
| dolphm | mdurnosvistov_: i was just leaving a comment on the bug | 18:36 |
| dolphm | morganfainberg: and yes! | 18:36 |
| dolphm | that'd be great | 18:36 |
| morganfainberg | dolphm, will check w/ infra and see if i did it all 100% right. | 18:36 |
| *** hartsocks has quit IRC | 18:36 | |
| dolphm | morganfainberg: THANK YOU! | 18:36 |
| morganfainberg | i followed the instructions...but well, it isn't clear about some things (since it's mostly directed at stackforge repos) | 18:37 |
| *** dkehn__ is now known as dkehn_ | 18:37 | |
| dolphm | mdurnosvistov_: https://bugs.launchpad.net/oslo/+bug/1266962/comments/26 | 18:38 |
| mdurnosvistov_ | dolphm: Yeah I saw comment =) Thanks! | 18:38 |
| uvirtbot | Launchpad bug 1266962 in oslo "Remove set_time_override in timeutils" [Low,In progress] | 18:38 |
| *** sandywalsh has quit IRC | 18:38 | |
| SergeyLukjanov | morganfainberg, I can help you new project creation | 18:39 |
| SergeyLukjanov | morganfainberg, and note that there are some problems now with manage-projects that blocks creating new projects, so, you'll need to wait for some time before it'll be fixed | 18:39 |
| dolphm | SergeyLukjanov: is there a fix in progress somewhere? or a bug report? | 18:39 |
| morganfainberg | SergeyLukjanov, good to know. | 18:40 |
| SergeyLukjanov | dolphm, issue reportd, let me find the link | 18:40 |
| *** mlegault has joined #openstack-dev | 18:40 | |
| SergeyLukjanov | dolphm, morganfainberg, https://bugs.launchpad.net/openstack-ci/+bug/1242569 | 18:40 |
| uvirtbot | Launchpad bug 1242569 in openstack-ci "manage-projects error on new project creation" [Critical,In progress] | 18:40 |
| *** promulo has joined #openstack-dev | 18:40 | |
| *** xqueralt has joined #openstack-dev | 18:40 | |
| morganfainberg | SergeyLukjanov, this also is a bit different as it's not a stackforge project, it's [without changing] under keystone's repo, so it would need to mirror more closely keystone stuff. so i wanted to run it by infra as well. | 18:40 |
| dolphm | morganfainberg: good news! https://launchpad.net/kite is available | 18:41 |
| morganfainberg | dolphm, WOOT! | 18:41 |
| morganfainberg | dolphm, are we renaming it kite? | 18:41 |
| dolphm | morganfainberg: please | 18:41 |
| SergeyLukjanov | morganfainberg, you'd like to extract a part of existing repo to the separated repo, am I right? | 18:41 |
| morganfainberg | dolphm, done. | 18:41 |
| morganfainberg | SergeyLukjanov, the way i see it is we create openstack/kite repo | 18:41 |
| promulo | hi guys, anyone available to help a devstack newbie? :) | 18:41 |
| morganfainberg | SergeyLukjanov, then we extract the code and add it to that new repo | 18:42 |
| morganfainberg | SergeyLukjanov, the code will need some restructuring in either case. | 18:42 |
| dolphm | morganfainberg: and at any point, i'd be happy to s/kds/kite/g | 18:42 |
| *** arnaud has joined #openstack-dev | 18:42 | |
| *** arnaud__ has joined #openstack-dev | 18:42 | |
| morganfainberg | dolphm, i'll make the repo openstack/kite in my commit no problem | 18:42 |
| SergeyLukjanov | morganfainberg, you'd probably like to create a repo on github and use git filter-branch to extract needed code with git history | 18:42 |
| morganfainberg | SergeyLukjanov, hum. | 18:42 |
| SergeyLukjanov | morganfainberg, and then fix some stuff like renaming | 18:42 |
| *** rossella_s has quit IRC | 18:42 | |
| *** luisbg has joined #openstack-dev | 18:43 | |
| morganfainberg | SergeyLukjanov, not sure how well that will work. the code is very very intertwined with keystone's atm | 18:43 |
| SergeyLukjanov | and import a gerrit project from this repo | 18:43 |
| *** tjones has quit IRC | 18:43 | |
| morganfainberg | SergeyLukjanov, I can try that though | 18:43 |
| morganfainberg | SergeyLukjanov, but in either case, need to do the repo in infra | 18:43 |
| SergeyLukjanov | morganfainberg, oh, it works good only for separated folders ;( | 18:43 |
| morganfainberg | SergeyLukjanov, yeah some of the code is separated. | 18:43 |
| *** ijw has joined #openstack-dev | 18:44 | |
| morganfainberg | SergeyLukjanov, i'll work on that bit in parallel | 18:44 |
| SergeyLukjanov | morganfainberg, sure, the only option is too clone it from keystone or from the already prepared project | 18:44 |
| *** tjones has joined #openstack-dev | 18:44 | |
| morganfainberg | dolphm, also https://review.openstack.org/#/c/72808/ ^_^ | 18:44 |
| morganfainberg | dolphm, but not sure if we want that during icehouse. | 18:44 |
| luisbg | https://wiki.openstack.org/wiki/HowToContribute mentions this channel, but not in which network it is. should I edit it so it does? | 18:44 |
| dolphm | morganfainberg: oh cool | 18:45 |
| dolphm | morganfainberg: that'd be great! | 18:45 |
| morganfainberg | SergeyLukjanov, ok, so it isn't a good idea to extract the history and then replay it into a current repo? | 18:45 |
| morganfainberg | SergeyLukjanov, better to have a repo that can be used as the base? i'm confused | 18:45 |
| dolphm | luisbg: i'd suggest linking to https://wiki.openstack.org/wiki/IRC | 18:46 |
| luisbg | dolphm, doing so :) | 18:46 |
| SergeyLukjanov | morganfainberg, when the project in gerrit will be created, you'll have no permissions to push code to repo directly | 18:46 |
| *** denis_ has quit IRC | 18:47 | |
| SergeyLukjanov | to fix history for example | 18:47 |
| SergeyLukjanov | so, it's much easier to add if you will do it in own repo | 18:47 |
| *** AlexF has quit IRC | 18:47 | |
| morganfainberg | SergeyLukjanov, and then have infra fork that repo? | 18:47 |
| SergeyLukjanov | morganfainberg, yup | 18:47 |
| alex-foo | can someone please check https://review.openstack.org/#/c/72751/ - i can't get it to pass the tests | 18:48 |
| morganfainberg | SergeyLukjanov, ah. | 18:48 |
| luisbg | dolphm, done | 18:48 |
| *** denis_cavalcante has joined #openstack-dev | 18:48 | |
| SergeyLukjanov | morganfainberg, IMO it'll be just easier ;) | 18:48 |
| *** jprovazn has joined #openstack-dev | 18:48 | |
| dolphm | morganfainberg: i've been defeated. not sure how to proceed http://i.imgur.com/wnW6WFC.png | 18:48 |
| dolphm | morganfainberg: there's nothing at that URL... | 18:48 |
| *** neelashah has joined #openstack-dev | 18:49 | |
| morganfainberg | SergeyLukjanov, ok and what is the way to communicate the fork need to infra? | 18:49 |
| *** Alexei_987 has quit IRC | 18:49 | |
| morganfainberg | dolphm, huh. | 18:49 |
| morganfainberg | dolphm, wtf? | 18:50 |
| morganfainberg | dolphm, *goes and looks* | 18:50 |
| termie | dolphm: is justinsb ever online anymore? | 18:50 |
| SergeyLukjanov | morganfainberg, hm, what do you mean? | 18:50 |
| dolphm | termie: not active, if he is | 18:50 |
| dolphm | termie: at least not in this channel | 18:50 |
| termie | dolphm: righto | 18:50 |
| *** zzelle has quit IRC | 18:51 | |
| *** saju_m has quit IRC | 18:51 | |
| morganfainberg | SergeyLukjanov, ok, so if i have a repo on say github, what do i do to communicate that it is the one to be forked from? | 18:51 |
| *** xqueralt has quit IRC | 18:51 | |
| *** tjones has quit IRC | 18:51 | |
| morganfainberg | SergeyLukjanov, because afaict the infra commit just creates the repo and the gerrit stuff, doesn't do anything to populate the repo | 18:51 |
| morganfainberg | SergeyLukjanov, or is that just something that happens side-band? | 18:52 |
| SergeyLukjanov | morganfainberg, you should create a config update like when creating new stackforge project and here you have an upstream propoerty | 18:52 |
| morganfainberg | SergeyLukjanov, ahhhhhh ok | 18:52 |
| *** alop has quit IRC | 18:52 | |
| SergeyLukjanov | morganfainberg, it links to the repo that'll be the base for gerrit project | 18:52 |
| dolphm | morganfainberg: what's the upstream? keystone? | 18:52 |
| morganfainberg | SergeyLukjanov, i see | 18:52 |
| *** eglynn-afk has joined #openstack-dev | 18:53 | |
| *** kgriffs_afk is now known as kgriffs | 18:53 | |
| SergeyLukjanov | dolphm, two options - (upstream==keystone and then cleanup with super large commit => bad history) vs. (upstream==some_prepared_repo with cleanup done => good git history good be done) | 18:53 |
| morganfainberg | dolphm, i think SergeyLukjanov is saying we should create a new repo and clean it up | 18:53 |
| SergeyLukjanov | morganfainberg, dolphm, that's your choise | 18:54 |
| morganfainberg | dolphm, for clean history | 18:54 |
| *** Tross has quit IRC | 18:54 | |
| morganfainberg | SergeyLukjanov, option 3, lose history, 1 import to clean repo | 18:54 |
| morganfainberg | dolphm, ^ | 18:54 |
| SergeyLukjanov | morganfainberg, exactly | 18:54 |
| SergeyLukjanov | option 4 is to squash keystone history to one commit ;) | 18:54 |
| dolphm | i vote option 3, starting with the current state | 18:54 |
| morganfainberg | dolphm, that was my thought | 18:54 |
| SergeyLukjanov | and I hope I can generate more crazy ideas :) | 18:55 |
| dolphm | delete all the keystone stuff though (an intermediary prep commit that is never tracked) | 18:55 |
| morganfainberg | SergeyLukjanov, yeah we'll extract the KDS stuff and squash it into a single commit, commit #1 in the new repo | 18:55 |
| morganfainberg | dolphm, ++ | 18:55 |
| dolphm | morganfainberg: i'm glad that made sense | 18:55 |
| morganfainberg | dolphm that was my initial thought | 18:55 |
| *** tjones has joined #openstack-dev | 18:55 | |
| morganfainberg | SergeyLukjanov, ok, so we do no upstream since this should be an empty repo | 18:55 |
| *** markmcclain has quit IRC | 18:55 | |
| morganfainberg | SergeyLukjanov, based upon how we're trying to split this | 18:56 |
| morganfainberg | dolphm, i was toying with a strange idea.... | 18:56 |
| *** jamespage_ has joined #openstack-dev | 18:56 | |
| SergeyLukjanov | morganfainberg, that's your choice ;) | 18:57 |
| *** markmcclain has joined #openstack-dev | 18:57 | |
| morganfainberg | dolphm, eh, we can talk about my crazy idea later... it has too many broad implications that i don't want to exlpain outside of keystone folks atm ;) | 18:57 |
| morganfainberg | dolphm, nevermind! will bring it up once this is all split out and working happily | 18:58 |
| dolphm | morganfainberg: what's the one line summary? | 18:58 |
| dolphm | morganfainberg: 140 chars or less | 18:58 |
| *** Gordonz has joined #openstack-dev | 18:58 | |
| *** Gordonz has quit IRC | 18:58 | |
| morganfainberg | dolphm, Identity Program, 2 projects, able to support separate core teams. | 18:58 |
| *** yolanda has quit IRC | 18:58 | |
| morganfainberg | dolphm, all under Identity program, and identity PTL (you) | 18:58 |
| *** jay-lau-513 has quit IRC | 18:59 | |
| *** Gordonz has joined #openstack-dev | 18:59 | |
| dolphm | morganfainberg: that's not entirely crazy -- that's sort of been suggested on some mailing list (TC?) | 18:59 |
| *** jay-lau-513 has joined #openstack-dev | 18:59 | |
| morganfainberg | dolphm, hm.. i could make kite this. | 18:59 |
| morganfainberg | dolphm, you know what, i'll go ahead and try and submit that idea with this repo | 18:59 |
| *** ndipanov has joined #openstack-dev | 19:00 | |
| morganfainberg | dolphm, whats the worst thing that happens, i'm told just use the keystone-* stuff we have. | 19:00 |
| *** saju_m has joined #openstack-dev | 19:01 | |
| morganfainberg | dolphm, LP project name: kite-key-distribution? | 19:01 |
| morganfainberg | dolphm, since 'kite' seems to be taken | 19:01 |
| dolphm | morganfainberg: i was thinking about opening a bug against lp | 19:01 |
| dolphm | maybe it's a long-inactive project or something | 19:01 |
| *** jamespage_ has quit IRC | 19:01 | |
| morganfainberg | dolphm, ++ that works too | 19:01 |
| morganfainberg | dolphm, :) | 19:01 |
| dolphm | morganfainberg: do you need a LP link right now? | 19:01 |
| *** pmathews has joined #openstack-dev | 19:02 | |
| *** joesavak has joined #openstack-dev | 19:02 | |
| *** ykhodork has joined #openstack-dev | 19:02 | |
| *** sahid has quit IRC | 19:02 | |
| *** sn6i23a has joined #openstack-dev | 19:03 | |
| dolphm | morganfainberg: and openstack/keydist-api/ or is that still /identity-api/? | 19:03 |
| *** kgriffs is now known as kgriffs_afk | 19:03 | |
| *** julienvey_ has quit IRC | 19:03 | |
| morganfainberg | dolphm, identity-api imo | 19:03 |
| dolphm | morganfainberg: p.s. i can never spell distrobution correctly | 19:04 |
| morganfainberg | dolphm, it's "identity program" | 19:04 |
| morganfainberg | so it would still be identity-api | 19:04 |
| *** chuck_ is now known as zul | 19:04 | |
| *** zul has quit IRC | 19:04 | |
| *** zul has joined #openstack-dev | 19:04 | |
| morganfainberg | just liek if barbican had become part of identity, i'd expect it to be part of identity-api | 19:04 |
| morganfainberg | though that brings up a bigger issue | 19:04 |
| dolphm | k; then the .md file would top-level against identity-api-v3.md ? keydist-api-v1.md ? | 19:04 |
| marun | ewindisch: ping | 19:04 |
| morganfainberg | maybe it should be kdist-api repo | 19:04 |
| *** jsavak has quit IRC | 19:05 | |
| morganfainberg | dolphm, since the keydist core would control that vs. keystone-core | 19:05 |
| *** martyntaylor has left #openstack-dev | 19:05 | |
| morganfainberg | dolphm, yeah i think we need a separate api repo too | 19:05 |
| morganfainberg | =/ | 19:05 |
| dolphm | morganfainberg: is that a big deal? | 19:05 |
| *** eglynn-afk has quit IRC | 19:05 | |
| *** e0ne has joined #openstack-dev | 19:06 | |
| morganfainberg | dolphm, nah | 19:06 |
| *** henrynash has quit IRC | 19:06 | |
| *** colinmcnamara has quit IRC | 19:06 | |
| sdague | jd__: so this will be going into all the runs, but it's especially interesting how often ceilometer-collector is the most expensive process on the test nodes - http://logs.openstack.org/86/72786/1/check/check-tempest-dsvm-full/21fbccd/logs/screen-dstat.txt.gz | 19:07 |
| morganfainberg | dolphm, so ok, we need a new keydist-api repo and a kite repo in my commit, nbd, good to know so i can add it all in | 19:07 |
| *** sarob has joined #openstack-dev | 19:07 | |
| morganfainberg | dolphm, description of kds: A service to store and distribute symmetrical keys under the OpenStack Identity program. | 19:10 |
| morganfainberg | dolphm ? | 19:10 |
| *** zul has quit IRC | 19:13 | |
| *** sarob has quit IRC | 19:13 | |
| *** zul has joined #openstack-dev | 19:14 | |
| *** Tross has joined #openstack-dev | 19:14 | |
| *** jckasper has joined #openstack-dev | 19:16 | |
| dolphm | morganfainberg: https://answers.launchpad.net/launchpad/+question/243837 | 19:16 |
| *** im_not_sure_why has joined #openstack-dev | 19:16 | |
| dolphm | morganfainberg: description sounds good | 19:16 |
| morganfainberg | dolphm, looks good to me | 19:16 |
| *** im_not_sure_why is now known as jmching | 19:16 | |
| dolphm | morganfainberg: i can propose to openstack/governance too | 19:16 |
| morganfainberg | dolphm, cool | 19:17 |
| morganfainberg | dolphm, ok i'm going to submit 2 reviews, 1 is the main repo, 2 will be keydist-api repo. | 19:18 |
| morganfainberg | dolphm, because i want to make sure i get both 100% correct and the api ones have doc building / publishing considerations i need to understand better | 19:19 |
| morganfainberg | dolphm, maaaaybe annegentle can explain it all to me ;) | 19:19 |
| dolphm | morganfainberg: openstack/python-kiteclient ? | 19:19 |
| *** Tross has quit IRC | 19:19 | |
| annegentle | morganfainberg: woahhh backscroll :) | 19:19 |
| morganfainberg | dolphm, ah, ok so client and server, that can be the first patchset. ++ | 19:19 |
| annegentle | morganfainberg: reading | 19:19 |
| morganfainberg | annegentle, mostly concerning adding a new -api repo with publishing | 19:20 |
| morganfainberg | annegentle, we want to splite the KDS service out from the keystone repo | 19:20 |
| *** rraja has joined #openstack-dev | 19:20 | |
| annegentle | morganfainberg: so, for networking related apis like vpnaas and so on, they use the netconn-api repo | 19:20 |
| morganfainberg | annegentle, still under identity program, but it's ugly to have 2 servers with disparate code (and tests) in one repo... and some things don't play nice e.g. auto-conf generation etc | 19:21 |
| morganfainberg | annegentle, so it might stay under identity-api? | 19:21 |
| annegentle | morganfainberg: and until about last week we were going to move all api doc specs into the <project> | 19:21 |
| annegentle | morganfainberg: so ideally no more repos please | 19:21 |
| morganfainberg | annegentle, Ah ok np, wont do another api repo :) | 19:21 |
| morganfainberg | annegentle, see this is why i ask :) | 19:21 |
| promulo | hey guys, did any of you ever enable healthnmon on a devstack (havana) installation? | 19:21 |
| morganfainberg | annegentle, answer makes this much simpler | 19:21 |
| morganfainberg | annegentle, thank you :) | 19:22 |
| annegentle | morganfainberg: good-o | 19:22 |
| annegentle | morganfainberg: I still think your markdown might let you down eventually as it's not easily cross linked but cross that bridge later | 19:22 |
| *** Mandell has quit IRC | 19:22 | |
| morganfainberg | annegentle, that is a totally different conversation :) | 19:22 |
| *** yeylon__ has joined #openstack-dev | 19:23 | |
| annegentle | morganfainberg: yep, just write all the things for now | 19:23 |
| *** jaybuff has joined #openstack-dev | 19:24 | |
| dolphm | morganfainberg: annegentle: openstack/governance proposal https://review.openstack.org/#/c/73022/ | 19:24 |
| *** jmching has quit IRC | 19:24 | |
| YorikSar | alex-foo: We're trying to land 2 changes that should make git-review tests more stable: https://review.openstack.org/70685 and https://review.openstack.org/71223 | 19:24 |
| dolphm | annegentle: do i need to do anything more than that? | 19:24 |
| morganfainberg | dolphm, :) | 19:24 |
| *** jaybuff has left #openstack-dev | 19:24 | |
| dolphm | annegentle: (for TC impact) | 19:24 |
| *** MaxV has joined #openstack-dev | 19:24 | |
| annegentle | dolphm: reading | 19:24 |
| *** vijendar has left #openstack-dev | 19:25 | |
| *** zzelle has joined #openstack-dev | 19:25 | |
| annegentle | dolphm: yeah that'll get the conversation going, might still request that ttx add it to the next TC meeting for any discussion, but a review on the /governance repo usually gets people thinking | 19:25 |
| *** MaxV has quit IRC | 19:25 | |
| dolphm | annegentle: that's my goal | 19:25 |
| YorikSar | alex-foo: I think it's easier for you to wait for them so that you won't have to recheck your change again and again. | 19:26 |
| *** vijendar has joined #openstack-dev | 19:26 | |
| *** kgriffs_afk is now known as kgriffs | 19:26 | |
| annegentle | dolphm: to me it's intriguing that you're doing kite rather than waiting for barbican? | 19:26 |
| *** IanGovett1 has joined #openstack-dev | 19:26 | |
| *** rraja has quit IRC | 19:26 | |
| morganfainberg | annegentle, lots of discussion and it's about getting secure messaging in sooner | 19:26 |
| morganfainberg | annegentle, and that barbican even in the short term isn't structured to handle the kds symmetrical keys needed in this fashion | 19:26 |
| morganfainberg | annegentle, eventually it should move under barbican, or another program (or barbican moves elsewhere) | 19:27 |
| morganfainberg | depending on incubation status etc | 19:27 |
| annegentle | morganfainberg: ok that's super helpful | 19:27 |
| *** thomasem has joined #openstack-dev | 19:27 | |
| *** IanGovett has quit IRC | 19:27 | |
| morganfainberg | but, afaict, kds is absolutely a form of identity, and core to services identifiying other services for messages on the bus | 19:28 |
| morganfainberg | it can be re-homed when there is a better place for it | 19:28 |
| *** yolanda has joined #openstack-dev | 19:28 | |
| dstanek | morganfainberg: are you still working on making the tests run in parallel? | 19:29 |
| morganfainberg | dstanek, it's a slow slog, but yes | 19:29 |
| morganfainberg | dstanek, it's blocking on some oslo syncs atm, | 19:29 |
| dstanek | morganfainberg: are you naming the db files by pid in your patch? | 19:29 |
| morganfainberg | dstanek, i will be | 19:29 |
| *** Mandell has joined #openstack-dev | 19:30 | |
| morganfainberg | dstanek, or.. i can use a tmpfile | 19:30 |
| morganfainberg | dstanek, actually... i might do that | 19:30 |
| *** dims has quit IRC | 19:30 | |
| dstanek | morganfainberg: hmmmm....checkout https://review.openstack.org/#/c/73024/1 | 19:30 |
| dstanek | maybe i should fix that to do it | 19:30 |
| *** nmagnezi has joined #openstack-dev | 19:30 | |
| *** tsekiyama has joined #openstack-dev | 19:30 | |
| morganfainberg | dstanek, i have some of the work for that in https://review.openstack.org/#/c/65870/ | 19:31 |
| ewindisch | marun: pong | 19:31 |
| *** colinmcnamara has joined #openstack-dev | 19:32 | |
| morganfainberg | dstanek, https://review.openstack.org/#/c/65870/ does most of it. but becasue we call load_backends multiple times in some tests, setup occurs in odd places and we end up with mixed backend data e.g. some setup data in KVS some in SQL | 19:32 |
| dolphm | annegentle: kite and barbican have slightly different scopes as well; kite can eventually use barbican | 19:32 |
| morganfainberg | dstanek, when we have a single test process we don't suffer test isolation issues. | 19:32 |
| marun | ewindisch: I was looking at the patches you have in flight for docker | 19:33 |
| *** bauzas has joined #openstack-dev | 19:33 | |
| *** pixelb has quit IRC | 19:33 | |
| morganfainberg | dstanek, which appear in the form i just described, some setup data in the wrong place... its ugly. | 19:33 |
| marun | ewindisch: And wondering whether your patches will ensure a working devstack+docker | 19:33 |
| ewindisch | marun: approaching... | 19:33 |
| marun | ewindisch: I've been broken since last week due to that glance issue :( | 19:33 |
| *** dims has joined #openstack-dev | 19:34 | |
| ewindisch | marun: glance issue? | 19:34 |
| morganfainberg | dstanek, feel free to steal what i did in that patch and/or co-opt it if you can resolve the tests. the bigger issue is that it makes the keystone tests fail in odd ways sometimes / sometimes not. so we'd potentially get transient failures | 19:34 |
| marun | ewindisch: https://github.com/dotcloud/docker-registry/issues/234 | 19:34 |
| morganfainberg | dstanek, and i think if we can ensure we call load_backends 1 time (raise an exception if we don't, or call it more than once) it'll resolve that transitent failure issue | 19:34 |
| dstanek | morganfainberg: i'll take a look; would is be better to split that up in to smaller patches to see what causes the breakage? | 19:35 |
| morganfainberg | dstanek, locally that failure case took me 30-40 test runs to duplicate, and then it failed 10 in a row, then not again. no changes. | 19:35 |
| ewindisch | marun: wonderful. I actually didn't see that. | 19:35 |
| morganfainberg | dstanek, i think the best bet is to slog slowly towards clean tests - the config fixture i have up is the latest in the chain (waiting on oslo sync of the new config fixture) | 19:36 |
| *** danielbruno has joined #openstack-dev | 19:36 | |
| marun | ewindisch: I swear it was working a couple of weeks ago, and now it's not possible to upload an image to the glance backend | 19:36 |
| marun | ewindisch: does your switch to uploading to glance directly bypass that? | 19:36 |
| *** yjiang51 has quit IRC | 19:36 | |
| morganfainberg | dstanek, then we can move to fixing ldap base and restructuring the tests that to load_backends multiple times or explicit reload_backends | 19:36 |
| ewindisch | marun: uploading to glance directly doesn't really work. | 19:37 |
| morganfainberg | dstanek, then i think we can fix load_backends to not be called more than once, then the parallel fix | 19:37 |
| marun | ewindisch: :( | 19:37 |
| ewindisch | marun: because glance doesn't know about the tags | 19:37 |
| morganfainberg | dstanek, about 4 patches from being there. | 19:37 |
| marun | ewindisch: ah, fair enough | 19:37 |
| ewindisch | marun: but you could upload something to glance and locally tag an image to the same name... that would probably work | 19:37 |
| morganfainberg | dstanek, couple that with Icehouse timelines and it becomes a very slow slog (esp. with new tests being added) | 19:37 |
| ewindisch | marun: (as a really ugly workaround) | 19:38 |
| marun | ewindisch: I'm a bit confused as to the relationship between the registry and glance | 19:38 |
| *** pmathews has quit IRC | 19:38 | |
| marun | ewindisch: if an image is uploaded directly, is the registry even needed? | 19:38 |
| ewindisch | marun: the registry can use glance as a storage background. It reads and writes to glance as if it were a filesystem storing its images. | 19:38 |
| marun | ewindisch: (docker newb, sorry) | 19:38 |
| *** novas0x2a|laptop has joined #openstack-dev | 19:39 | |
| *** krtaylor has quit IRC | 19:39 | |
| ewindisch | marun: but glance doesn't store tags, so the docker registry (currently) needs to store the tags somewhere else | 19:39 |
| ewindisch | potentially we could have glance store the tags and have the registry pull them correctly... | 19:39 |
| marun | ewindisch: ah, so uploading an image directly to glance bypasses the need to upload through the registry but tag maintenance by the registry is still required (for now)? | 19:39 |
| ewindisch | marun: well, you can't do tag maintenance in the registry if the image exists in glance. | 19:40 |
| marun | ewindisch: clear as mud :) | 19:40 |
| ewindisch | marun: so you basically need to upload the image through the registry (currently) | 19:40 |
| ewindisch | OR have an image in glance and a tag in your local docker database (not on the registry) | 19:40 |
| marun | ewindisch: ah, that's what I was missing | 19:41 |
| *** byeager has joined #openstack-dev | 19:41 | |
| marun | ewindisch: so is the registry unnecessary in the latter scheme? | 19:41 |
| *** gnorth has joined #openstack-dev | 19:41 | |
| ewindisch | marun: you might need a small patch for that to be true, but technically. | 19:42 |
| ewindisch | without the registry you can't make multiple hosts work. | 19:42 |
| marun | ewindisch: ah, ok | 19:42 |
| marun | ewindisch: so your patchs for bp/docker-glance-uploads is for single-host devstack then? | 19:43 |
| *** baoli has quit IRC | 19:43 | |
| ewindisch | marun: there is a reason that is a work-in-progress patch :) | 19:43 |
| marun | ewindisch: Apologies, just trying to understand | 19:44 |
| ewindisch | marun: but yes, the current state of the patch would probably work in a single-host devstack | 19:44 |
| marun | ewindisch: ok | 19:45 |
| ewindisch | marun: btw have you seen dockenstack? https://github.com/ewindisch/dockenstack | 19:45 |
| marun | huh | 19:46 |
| marun | ewindisch: that's pretty neat, though I'm not sure it will be useful for me | 19:47 |
| ewindisch | marun: it's intended for testing and development | 19:47 |
| marun | ewindisch: afaik, it's not possible to run ovs in a docker container is it? | 19:47 |
| *** eglynn-afk has joined #openstack-dev | 19:47 | |
| marun | ewindisch: (disclusure - my focus is networking/neutron) | 19:47 |
| ewindisch | marun: well, we're using a privileged container there, so it probably would work. | 19:48 |
| marun | ewindisch: erm | 19:48 |
| ewindisch | marun: (or could be made to work) | 19:48 |
| *** saju_m has quit IRC | 19:48 | |
| *** novas0x2a|laptop has quit IRC | 19:49 | |
| marun | ewindisch: I'd certainly like that to be the case. | 19:49 |
| marun | ewindisch: I'll have to find an expert to verify it, though. | 19:49 |
| ewindisch | that is to say, none of the linux capabilities are disabled. You can mknod files and read/write to any device you might like. | 19:49 |
| *** julienvey_ has joined #openstack-dev | 19:51 | |
| marun | ewindisch: I think the problem may be that ovs doesn't directly support the concept of containers | 19:52 |
| bknudson | things are finally starting to merge in keystone. | 19:52 |
| ewindisch | marun: that doesn't matter. | 19:52 |
| marun | ewindisch: It has a database that is per-host and tied into a kernel module | 19:52 |
| ewindisch | marun: well, you might have trouble using ovs inside multiple containers on a single host. | 19:52 |
| morganfainberg | bknudson, i noticed | 19:52 |
| ewindisch | marun: but in a single container, I wouldn't expect troubles | 19:53 |
| *** julienvey_ has quit IRC | 19:53 | |
| marun | ewindisch: uh | 19:53 |
| *** geraint2 has joined #openstack-dev | 19:53 | |
| marun | ewindisch: a single container + host => 2 | 19:53 |
| *** colinmcnamara has quit IRC | 19:53 | |
| ewindisch | marun: single container + host = 1 :) | 19:54 |
| marun | ewindisch: sure, if you say so. | 19:54 |
| ewindisch | marun: one copy of the user-space tools running in the container. No user-space tools running on the host | 19:54 |
| *** Alexei_987 has joined #openstack-dev | 19:54 | |
| *** gnorth has quit IRC | 19:54 | |
| lifeless | one kernel | 19:54 |
| marun | ewindisch: if you say so. | 19:54 |
| lifeless | one ovs :) | 19:54 |
| ewindisch | right | 19:54 |
| marun | lifeless: do you have any experience running ovs in a container? | 19:55 |
| *** moted has quit IRC | 19:55 | |
| lifeless | marun: a little, indirectly | 19:55 |
| ewindisch | marun: it's more about "managing ovs from the container" and letting the userspace tools access any devices or syscalls it might need | 19:55 |
| lifeless | marun: we didnt fully explore it because block devices were not namespaced and we couldn't use iscsi | 19:56 |
| lifeless | from within the container | 19:56 |
| marun | ewindisch: right, which implies that we don't really have true isolation | 19:56 |
| lifeless | marun: but, I'd glue the tools to the kernel via a remote manager, not local. | 19:56 |
| marun | ewindisch: the container just happens to be the only client of ovs | 19:56 |
| ewindisch | marun: right, precisely. There is no isolation with a privileged container | 19:56 |
| *** tjones has quit IRC | 19:57 | |
| lifeless | thereis no privilege within an isolated container? | 19:57 |
| lifeless | </humour> | 19:57 |
| *** rfolco has quit IRC | 19:57 | |
| *** avishay_away is now known as avishay | 19:57 | |
| *** tjones has joined #openstack-dev | 19:57 | |
| ewindisch | privileged containers just means that there are no linux capabilities dropped | 19:57 |
| *** gordc has joined #openstack-dev | 19:58 | |
| marun | ewindisch: I'm afraid I'm a little fuzzy on the benefits of openstack in a container. Do you have specific use-cases in mind? | 19:59 |
| *** markmc has joined #openstack-dev | 19:59 | |
| marun | ewindisch: I mean, for specific services the benefits of containerization are pretty clear. | 19:59 |
| ewindisch | marun: my use-case is for ease and speed of launching it for dev and testing | 19:59 |
| *** avishay is now known as avishay_away | 20:00 | |
| ewindisch | containerizing specific services is more practical for production use-cases | 20:00 |
| ewindisch | marun: for instance, dockenstack runs as a CI job: http://openstack-ci.docker.io/job/docker-tempest-full-gerrit-trigger/145/console | 20:01 |
| *** tjones has quit IRC | 20:01 | |
| *** geraint2 has quit IRC | 20:01 | |
| *** gnorth has joined #openstack-dev | 20:01 | |
| alex-foo | YorikSar: is the port determinism thing causing the hanging? | 20:02 |
| marun | ewindisch: so it's a way of providing a clean state faster than vm boot? | 20:02 |
| *** zul has quit IRC | 20:03 | |
| marun | ewindisch: is there any advantage in how long it takes to deploy? | 20:03 |
| ewindisch | marun: basically, as well as making it easier to define that clean state | 20:03 |
| *** zul has joined #openstack-dev | 20:03 | |
| marun | ewindisch: hmmm… That does sound promising. | 20:03 |
| ewindisch | i.e. I can define that environment via a dockerfile instead of some more customized scripts | 20:03 |
| YorikSar | alex-foo: We're not sure. But port determinism should lower probability of error and timeout will ensure you don't need to wait an hour for another try. | 20:03 |
| *** denis_cavalcante has quit IRC | 20:04 | |
| marun | ewindisch: well, using vagrant+puppet to create a consistent clean slate works well for me | 20:04 |
| *** zul has quit IRC | 20:04 | |
| *** otherwiseguy has joined #openstack-dev | 20:04 | |
| marun | ewindisch: but it takes a couple of minutes vs what would presumably be seconds for docker | 20:04 |
| YorikSar | Unfortunately there's shortage of py26 nodes... So all our changes are waiting for them now. | 20:04 |
| *** ijw_ has joined #openstack-dev | 20:04 | |
| alex-foo | YorikSar: perhaps an admin can debug the hanging test with an strace/gdb and get to the bottom of it | 20:04 |
| marun | ewindisch: very interesting stuff | 20:05 |
| *** joesavak has quit IRC | 20:05 | |
| marun | ewindisch: Backing up a little, have you seen those issues with the docker registry? | 20:05 |
| YorikSar | alex-foo: Timeout should help with that too. We can collect as much data as we can and analyse it later. | 20:06 |
| ewindisch | marun: true, puppet or chef helps. I can still run a VM and have it start my docker container, but I'm not tied to starting a VM each time, either. | 20:06 |
| ewindisch | marun: the main issue for me is not being able to upload directly to Glance, or attempting to run with multiple registries. | 20:06 |
| marun | ewindisch: so if you do a clean devstack+docker run you end up with a working system? | 20:07 |
| ewindisch | the glance upload issue, I hadn't seen, but I hadn't checked/diagnosed it yet. | 20:07 |
| *** ijw has quit IRC | 20:07 | |
| *** mlavalle has joined #openstack-dev | 20:07 | |
| marun | ewindisch: I want to prototype neutron support for the docker driver (so that we can stress-test neutron against containers instead of vm's or the noop virt driver), but I can't get a working deployment :( | 20:08 |
| ewindisch | marun: I don't bring it up and test it manually and try to use it, I have it automated... and that automation isn't yet working. | 20:08 |
| ewindisch | the idea being that once the automation is working, everything should be nearly perfect out of the box. | 20:08 |
| marun | ewindisch: if wishes were horses ;) | 20:09 |
| bknudson | stevemar dolphm: I don't see any warnings generated during doc build now | 20:10 |
| ewindisch | marun: well, I'm certain that devstack+docker will be working out of the box once it passes tempest. | 20:10 |
| dolphm | bknudson: ooh yeah, those two changes should have just merged | 20:10 |
| dolphm | bknudson: rechecking https://review.openstack.org/#/c/72492/ | 20:10 |
| marun | ewindisch: well, it's non-functional out of the box right now due to that registry problem. | 20:11 |
| ewindisch | marun: but I agree that registry issues are the main issue. | 20:11 |
| ewindisch | *problem | 20:11 |
| marun | ewindisch: Is there any priority on fixing that? | 20:11 |
| bknudson | dolphm: the 72492 build has some failures due to kombu and lockfile... | 20:11 |
| marun | ewindisch: I'd be happy to help but I didn't want to duplicate effort. | 20:12 |
| ewindisch | marun: that's why I've shifted to having devstack pull the registry the index instead of from a static and infrequently updated S3 download. | 20:12 |
| bknudson | dolphm: I worked around that by .tox/docs/bin/pip install kombu lockfile | 20:12 |
| *** yolanda has quit IRC | 20:13 | |
| *** Tross has joined #openstack-dev | 20:13 | |
| ewindisch | marun: I have a priority to making docker+devstack work, so that includes any dependencies such as that registry bug. | 20:15 |
| ewindisch | marun: however, I'm not working on that bug at this very moment, so I'd appreciate any help that is offered. | 20:15 |
| marun | ewindisch: there is a launchpad bug that somebody marked invalid because it is an upstream issue: https://bugs.launchpad.net/devstack/+bug/1277852 | 20:17 |
| uvirtbot | Launchpad bug 1277852 in devstack "Docker images are broken in glance " [Undecided,Invalid] | 20:17 |
| marun | ewindisch: Would it make sense to reopen it to ensure it gets tracked for openstack? | 20:17 |
| marun | ewindisch: I'm not really clear on how you want openstack+docker bugs managed | 20:18 |
| ewindisch | marun: I'd rather it were all in openstack and we linked to upstream bugs where relevant to getting those openstack bugs closed. | 20:18 |
| marun | ewindisch: ok | 20:19 |
| ewindisch | marun: but there are certainly will be edge-case examples... | 20:19 |
| *** novas0x2a|laptop has joined #openstack-dev | 20:20 | |
| marun | ewindisch: I'll see if I can make headway. No promises given my inexperience with the subject matter. | 20:20 |
| ewindisch | thanks for bringing it to my attention. | 20:22 |
| *** pixelb has joined #openstack-dev | 20:22 | |
| *** kamalic2 has joined #openstack-dev | 20:23 | |
| *** mst89 has quit IRC | 20:24 | |
| *** sushils has joined #openstack-dev | 20:24 | |
| gnorth | Has anyone here worked on the Cinder TaskFlow stuff? I have a change that could do with an expert taking a quick look to see if I'm on the right track. | 20:24 |
| *** eglynn-afk has quit IRC | 20:28 | |
| *** tjones has joined #openstack-dev | 20:29 | |
| *** tjones has quit IRC | 20:30 | |
| *** tjones has joined #openstack-dev | 20:32 | |
| ihrachys | I have a patch for havana approved to be merged in the upcoming release: https://review.openstack.org/#/c/72754/ But it fails on the same UT (test_snapshot_pattern.TestSnapshotPattern) on isolated tasks with SSHTimeout. I wonder whether there is some bug in stable tempest (ssh.py?) that was fixed in master but didn't reach stable branch yet that could influence the difference between master neutron and | 20:32 |
| ihrachys | its stable branch? I see there are several patches for ssh.py in tempest master that are not part of tempest havana... Can it be a reason? | 20:32 |
| *** eharney has joined #openstack-dev | 20:32 | |
| *** promulo has quit IRC | 20:33 | |
| *** SumitNaiksatam has joined #openstack-dev | 20:33 | |
| *** eglynn-afk has joined #openstack-dev | 20:33 | |
| *** joesavak has joined #openstack-dev | 20:36 | |
| bknudson | morganfainberg: http://status.openstack.org/reviews/reviewday.json | 20:36 |
| morganfainberg | bknudson, woot | 20:36 |
| *** devoid1 has joined #openstack-dev | 20:37 | |
| morganfainberg | bknudson, now just need to wait for https://review.openstack.org/#/c/72815/ to be active | 20:37 |
| morganfainberg | bknudson, and it'll be super awesome | 20:37 |
| bknudson | morganfainberg: now if that were integrated with next-review somehow... | 20:37 |
| morganfainberg | bknudson, shouldn't be hard to do that | 20:37 |
| *** devoid has quit IRC | 20:38 | |
| jgriffith | gnorth: what are you looking at? | 20:38 |
| morganfainberg | dstanek, https://review.openstack.org/#/c/71024/ is covered by https://review.openstack.org/#/c/70523 not sure if you want to do both. | 20:39 |
| *** galstrom is now known as galstrom_zzz | 20:40 | |
| *** abhirc has joined #openstack-dev | 20:40 | |
| dstanek | morganfainberg: i do like my test :-) maybe i'll base mine on yours | 20:42 |
| morganfainberg | sure! | 20:42 |
| morganfainberg | dstanek, that much i can agree on. | 20:42 |
| *** krtaylor has joined #openstack-dev | 20:43 | |
| *** markmcclain has quit IRC | 20:43 | |
| morganfainberg | actually | 20:43 |
| morganfainberg | dstanek, should probaly just fix the last test down there https://review.openstack.org/#/c/70523/9/keystone/tests/test_kvs.py | 20:44 |
| morganfainberg | dstanek, i do like your test | 20:44 |
| *** joesavak has quit IRC | 20:45 | |
| *** mohits has quit IRC | 20:45 | |
| lifeless | morganfainberg: we do need to talk about the keystone recommended default setup though | 20:46 |
| lifeless | morganfainberg: because thats what we'll run | 20:46 |
| morganfainberg | lifeless, ++ yes | 20:46 |
| gnorth | jgriffith - it is my volume import stuff at https://review.openstack.org/#/c/72501/ | 20:46 |
| morganfainberg | lifeless, sure thing :) | 20:46 |
| lifeless | morganfainberg: and right now, - well that 30GB comment, was real data, not troll | 20:46 |
| morganfainberg | lifeless, i know :( i suffer from the same issue in production now | 20:46 |
| lifeless | morganfainberg: from a 40 node baremetal cloud, so rouhgly 2-3 virt hypervisor cloud size. | 20:47 |
| morganfainberg | lifeless, it's painful. | 20:47 |
| lifeless | morganfainberg: ah ok :) | 20:47 |
| gnorth | jgriffith - adds the new manage volume properties, I think I've done it right, but it seeemed too simple - not sure if I have to do anything else around managing persistence, rollback etc. I don't think so. | 20:47 |
| dolphm | morganfainberg: i hear you promising a patch for ephemeral tokens in #openstack-meeting -- link? | 20:47 |
| morganfainberg | dolphm, yes yes | 20:47 |
| * dolphm waits patiently | 20:47 | |
| *** DinaBelova is now known as DinaBelova_ | 20:47 | |
| morganfainberg | dolphm, i know... | 20:47 |
| morganfainberg | dolphm, i have my dev environment setup was working on it. doing quick review break for the low hanging fruit then on to that | 20:47 |
| morganfainberg | dolphm :) | 20:48 |
| *** harlowja is now known as harlowja_away | 20:48 | |
| morganfainberg | lifeless, we do an hourly cron atm to cleanup expired tokens, and keep the token TTL low(er than 86400) | 20:48 |
| dolphm | </harassment_by_ptl> | 20:48 |
| morganfainberg | lifeless, it helps. but if you have too many tokens the flush expired tokens causes massive gap lock prioblems...and hangs up keystone horribly | 20:48 |
| morganfainberg | lifeless, can discuss later though. | 20:48 |
| morganfainberg | dolphm, hehe | 20:49 |
| *** danielbruno has quit IRC | 20:49 | |
| gordc | markmc: ping | 20:50 |
| morganfainberg | dolphm, -2 https://review.openstack.org/#/c/70656/ no, just no. the whole concept of using something that could be valid data as an identitifier for "None" really rubs me the wrong way. | 20:51 |
| morganfainberg | dolphm, i know you went with -1 but, i think we need to rethink that (and perhaps have something more parse-friendly) before a change goes in. | 20:52 |
| morganfainberg | s/change/change like that/ | 20:52 |
| lifeless | morganfainberg: having an external cron job is a bit of a bug IMO; a periodic task running the background of the server would be better | 20:53 |
| stevemar | dolphm, we going ahead with: https://review.openstack.org/#/c/72089 ? | 20:53 |
| *** samuelqueiroz has joined #openstack-dev | 20:53 | |
| lifeless | morganfainberg: because it could increase its activity as load inreases | 20:53 |
| morganfainberg | lifeless, except keystone doesn't reimplement cron (like some projects). | 20:53 |
| lifeless | morganfainberg: and tune it's batch size to keep commit times low | 20:53 |
| *** arborism has joined #openstack-dev | 20:54 | |
| lifeless | morganfainberg: you don't need to reimplement cron: running arbitrary commands isn't the problem. | 20:54 |
| morganfainberg | lifeless, i know. | 20:54 |
| morganfainberg | lifeless, i agree a clean way for batch-size stuff would be good | 20:54 |
| *** browne has quit IRC | 20:54 | |
| morganfainberg | lifeless, and worth making part of the real solution™ | 20:54 |
| *** boris-42_ has quit IRC | 20:55 | |
| *** dripton has quit IRC | 20:55 | |
| *** browne has joined #openstack-dev | 20:55 | |
| lifeless | morganfainberg: LP has a nice auto tuning loop you could study for inspiration, but the basic idea is to have a batch size and count how long a delete takes. if its higher than (threshold), reduce the size (to some lower limit) | 20:55 |
| lifeless | if its lower than (threashold), increase the size (to some upper limit) | 20:56 |
| morganfainberg | lifeless, though, to be fair, i think the real solution is to never store tokens. | 20:56 |
| morganfainberg | lifeless, really push to the ephemeral token concept, and not need the persistence | 20:56 |
| morganfainberg | lifeless, but that is at least L before we'd be there with a removed persistence store | 20:56 |
| samuelqueiroz | Hi, I'm using Keystone and I'm creating a customized version of the policy | 20:57 |
| samuelqueiroz | I'd like to know if there is a way to verify that an optional filter (e.g. the domain filter when listing users) is being given | 20:57 |
| morganfainberg | lifeless assuming we deprecate SQL backend for tokens (and UUID) in J if all goes well. | 20:57 |
| *** arborism is now known as amcnr | 20:57 | |
| *** amcnr is now known as amcrn | 20:57 | |
| *** jdob has quit IRC | 20:58 | |
| *** d0ugal has quit IRC | 20:58 | |
| morganfainberg | lifeless, there is also the thought of where to run that thread/process, as keystone is designed to run multiple times under wsgi, or across multiple servers. it's def. worth thinking about. | 20:58 |
| *** jdob has joined #openstack-dev | 20:58 | |
| *** jamielennox|away is now known as jamielennox | 20:58 | |
| *** dripton has joined #openstack-dev | 20:58 | |
| *** SpamapS_ is now known as SpamapS | 20:58 | |
| *** SpamapS has quit IRC | 20:58 | |
| *** SpamapS has joined #openstack-dev | 20:58 | |
| *** boris-42_ has joined #openstack-dev | 20:59 | |
| *** zul has joined #openstack-dev | 20:59 | |
| *** yjiang51 has joined #openstack-dev | 21:00 | |
| lifeless | morganfainberg: just run it in each process :) | 21:01 |
| morganfainberg | lifeless, hehe | 21:01 |
| *** thomasem has quit IRC | 21:01 | |
| *** eglynn-afk is now known as eglynn | 21:01 | |
| *** asalkeld has quit IRC | 21:02 | |
| *** asalkeld has joined #openstack-dev | 21:02 | |
| morganfainberg | dolphm, for the disable notifications, we expect an "updated" and a "disabled" notification if we disable a project? | 21:02 |
| morganfainberg | dolphm, actually... i guess that makes sense. | 21:02 |
| *** jckasper_ has joined #openstack-dev | 21:03 | |
| *** raildo has quit IRC | 21:03 | |
| *** abhirc has quit IRC | 21:04 | |
| *** kgriffs is now known as kgriffs_afk | 21:04 | |
| *** jckasper has quit IRC | 21:04 | |
| *** kgriffs_afk is now known as kgriffs | 21:04 | |
| jgriffith | gnorth: interesting... | 21:05 |
| *** marekd is now known as marekd|away | 21:05 | |
| gnorth | jgriffith: good interesting? | 21:06 |
| *** jp_at_hp1 has quit IRC | 21:06 | |
| jgriffith | gnorth: sure... :) | 21:06 |
| jgriffith | gnorth: tieing in through the create_volume flow was something I hadn't thought about | 21:06 |
| jgriffith | gnorth: I mean... it didn't exist when I looked at this last :) | 21:07 |
| gnorth | Yeah, me neither, was one hell of a rebase. :-) | 21:07 |
| jgriffith | gnorth: yeah... sorry about that :) | 21:08 |
| *** jdob has quit IRC | 21:08 | |
| jgriffith | gnorth: especially since it was a moving target the last week or so | 21:08 |
| *** jruzicka has quit IRC | 21:08 | |
| *** jdob has joined #openstack-dev | 21:08 | |
| *** kamalic2 has quit IRC | 21:08 | |
| jgriffith | gnorth: so seems pretty sane, I'll pick at it here lately and run some tests on it | 21:08 |
| jgriffith | gnorth: seems well put together at first glance | 21:09 |
| *** denis_makogon_ has joined #openstack-dev | 21:09 | |
| *** baoli has joined #openstack-dev | 21:09 | |
| gnorth | jgriffith: It feels quite clean, I couldn't find a way to add the manage_volume as an action to the os-hosts API, which might be a little more REST-y. | 21:10 |
| *** samuelbercovici has joined #openstack-dev | 21:12 | |
| *** baoli has quit IRC | 21:12 | |
| *** hartsocks has joined #openstack-dev | 21:12 | |
| *** tjones has quit IRC | 21:13 | |
| *** topol has quit IRC | 21:13 | |
| *** samuelbercovici has quit IRC | 21:13 | |
| *** baoli has joined #openstack-dev | 21:13 | |
| *** derekh has joined #openstack-dev | 21:13 | |
| *** tjones has joined #openstack-dev | 21:14 | |
| *** rodrigods has joined #openstack-dev | 21:14 | |
| *** rodrigods has joined #openstack-dev | 21:14 | |
| *** pmathews has joined #openstack-dev | 21:15 | |
| *** ndipanov has quit IRC | 21:15 | |
| *** jprovazn has quit IRC | 21:15 | |
| *** mlegault has quit IRC | 21:16 | |
| stevemar | bknudson, apparently the patch to fix docs failed again, i took a peek at http://status.openstack.org/zuul/ for 72492 | 21:16 |
| bknudson | stevemar: looks like it doesn't like the missing lockfile package | 21:16 |
| jamielennox | dolphm, bknudson: am out for a bit but do you want to have a look over: https://wiki.openstack.org/wiki/VersionDiscovery i'll send it to -dev list later today for comments but feel free to polish up the wording or anything you see is missing | 21:17 |
| bknudson | stevemar: maybe can add it to test-requires.txt? | 21:17 |
| stevemar | or kombu | 21:17 |
| jamielennox | bknudson, dolphm: this will help: http://paste.openstack.org/show/64770/ | 21:18 |
| stevemar | bknudson, why not requirements? it's not used in test, it's part of oslo right | 21:21 |
| bknudson | stevemar: if you put it in requirements then that implies that it's required to even run the server | 21:21 |
| bknudson | and packagers will think that they have to include it | 21:21 |
| stevemar | bknudson, ahhh | 21:21 |
| bknudson | stevemar: actually, bnemec just posted something to the -dev mailing list... | 21:22 |
| *** tjones has quit IRC | 21:22 | |
| *** denis_makogon_ is now known as denis_makogon | 21:22 | |
| *** hartsocks has left #openstack-dev | 21:22 | |
| *** vartom1111111113 has joined #openstack-dev | 21:22 | |
| *** marcoemorais has quit IRC | 21:23 | |
| stevemar | bknudson, our mail servers have been going offline/online all day, very annoying. checking now | 21:23 |
| bknudson | stevemar: blame the mayor | 21:23 |
| bnemec | Oh good, so we're not the only ones running into that issue. :-) | 21:23 |
| *** yjiang51 has quit IRC | 21:23 | |
| stevemar | bknudson, can't blame him for everything | 21:23 |
| stevemar | bnemec, not at all | 21:23 |
| bknudson | bnemec: well... oslo is the cause of it for us | 21:24 |
| *** tzabal has left #openstack-dev | 21:24 | |
| *** mriedem has quit IRC | 21:24 | |
| *** martines has quit IRC | 21:24 | |
| bnemec | bknudson: Well, hopefully the wiki page Doug just sent to the list will help with that. | 21:25 |
| jgriffith | gnorth: personally I prefer the seperate extensions | 21:25 |
| *** markwash has quit IRC | 21:25 | |
| jgriffith | gnorth: ie the way you did it | 21:25 |
| bnemec | Get some of this stuff out of incubator and into libraries that have their own deps. | 21:25 |
| *** dprince has quit IRC | 21:26 | |
| *** pmathews has quit IRC | 21:26 | |
| bknudson | bnemec: the one we run into in keystone during doc build is kombu | 21:26 |
| bknudson | and also lockfile... although that seems like a legit requirement. | 21:26 |
| bknudson | but I think kombu is optional, depending on which rpc backend you use | 21:26 |
| morganfainberg | dolphm, yay! name is freed up! | 21:26 |
| bnemec | bknudson: Yeah, that's exactly the issue I'm hitting. | 21:27 |
| bnemec | kombu is included in most project's requirements, even though technically it isn't required. | 21:27 |
| bnemec | But qpid isn't, and there's understandable pushback to including it | 21:27 |
| bknudson | bnemec: since oauth is optional for keystone we only have it in test-requires.txt | 21:27 |
| *** nealph_ has joined #openstack-dev | 21:28 | |
| *** nplanel has quit IRC | 21:29 | |
| bnemec | bknudson: Yeah, I'm hoping our Python packaging gurus will chime in with a brilliant solution. :-) | 21:29 |
| *** markmc has quit IRC | 21:29 | |
| *** csd has joined #openstack-dev | 21:30 | |
| *** devoid1 is now known as devoid | 21:30 | |
| *** ayoung has joined #openstack-dev | 21:30 | |
| dhellmann | bnemec, bknudson: I don't think we'll ever reach a state where, for example, oslo.messaging lists an explicit dependency on all of the libraries used by all of the drivers | 21:31 |
| *** nplanel has joined #openstack-dev | 21:31 | |
| dhellmann | bnemec, bknudson: it might make sense to have some separate requirements files for when those drivers are in use, though, so an installer that knows how a configuration is being set up could pick the right file based on a driver or configuration name, and not have to keep up with the details that choice implies | 21:32 |
| *** colinmcnamara has joined #openstack-dev | 21:33 | |
| *** marcoemorais has joined #openstack-dev | 21:33 | |
| *** eglynn has quit IRC | 21:33 | |
| *** nealph_ has quit IRC | 21:33 | |
| *** nealph has joined #openstack-dev | 21:34 | |
| *** e0ne has quit IRC | 21:34 | |
| stevemar | bknudson, https://review.openstack.org/#/c/73072/ | 21:34 |
| bnemec | dhellmann: Yeah, that seems reasonable. I see oslo.messaging doesn't actually require any messaging backend. | 21:35 |
| dhellmann | bnemec: right, because if you're using qpid why should you install zmq? | 21:36 |
| bnemec | dhellmann: Or kombu, which most of the projects require right now. | 21:37 |
| dhellmann | bnemec: right | 21:38 |
| lifeless | dhellmann: so I think long term sure | 21:38 |
| lifeless | dhellmann: but I don't think we should have any default if we split it out | 21:38 |
| lifeless | dhellmann: e.g. if we test both pg and mysql, we should install either neither driver, or both. | 21:39 |
| lifeless | by dfault | 21:39 |
| lifeless | -or- | 21:39 |
| dhellmann | lifeless: that makes sense | 21:39 |
| lifeless | we should actually make a statement that backend X is *the backend* and everything else is /actually/ second class, not defacto second class | 21:39 |
| *** martines has joined #openstack-dev | 21:40 | |
| morganfainberg | dhellmann, btw, thank you very very much for the help eysterday :) | 21:41 |
| morganfainberg | got it all working | 21:41 |
| dhellmann | morganfainberg: excellent! | 21:42 |
| morganfainberg | lifeless do you know much about new repo creation (project) or should i bug another infra person to see if i have everything "correct | 21:43 |
| morganfainberg | " | 21:43 |
| *** promulo has joined #openstack-dev | 21:43 | |
| *** promulo has quit IRC | 21:43 | |
| morganfainberg | dhellmann, ^ you might have an answer as well, though i am unsure how close you are to infra at this point | 21:43 |
| *** promulo has joined #openstack-dev | 21:43 | |
| lifeless | morganfainberg: there's a guide | 21:44 |
| morganfainberg | lifeless, i followed it | 21:44 |
| *** tmclaugh[work] has quit IRC | 21:44 | |
| lifeless | morganfainberg: then you're all good ;) | 21:44 |
| morganfainberg | lifeless, well it's more along that i'm splitting code out of keystone's repo and about the [pypi] etc sections in zuul | 21:44 |
| morganfainberg | lifeless, i'll just poke people to make sure i trim down/modify what is there as needed. thanks! :) | 21:45 |
| *** dkehn__ has joined #openstack-dev | 21:46 | |
| dhellmann | morganfainberg: as it happens: https://wiki.openstack.org/wiki/Oslo/CreatingANewLibrary | 21:46 |
| morganfainberg | dhellmann, hehe | 21:47 |
| dhellmann | morganfainberg: that is a WIP, so let me know if you find issues | 21:47 |
| dhellmann | morganfainberg: SergeyLukjanov has been helping, and can answer questions, too | 21:47 |
| morganfainberg | dhellmannh, yeah SergeyLukjanov has been awesome | 21:47 |
| morganfainberg | here is my first pass https://review.openstack.org/#/c/73074/ for the zuul etc bits | 21:47 |
| morganfainberg | this isn't so much a lib as another project being split out form the keystone repo | 21:47 |
| morganfainberg | KDS | 21:47 |
| *** dkehn__ has quit IRC | 21:48 | |
| dhellmann | morganfainberg: you might be interested in the graduate.sh script in oslo-incubator -- it knows how to pull files out of git and preserve their history | 21:48 |
| *** alop has joined #openstack-dev | 21:48 | |
| dhellmann | dhellmann: I wrote the script, but markmc worked out the commands | 21:48 |
| *** dkehn__ has joined #openstack-dev | 21:48 | |
| morganfainberg | dhellmann, cool. i'll check that out | 21:48 |
| * dhellmann is now talking to himself | 21:48 | |
| ayoung | morganfainberg, https://blueprints.launchpad.net/python-keystoneclient/+spec/validate-service-endpoint | 21:48 |
| *** MaxV has joined #openstack-dev | 21:49 | |
| *** dkehn_ has quit IRC | 21:49 | |
| ayoung | morganfainberg, the author is working on that.... | 21:49 |
| morganfainberg | dhellmann, we might just collapse out the history for the first rev in a new repo. because of moving target issues | 21:49 |
| *** cagrev_ has quit IRC | 21:49 | |
| bknudson | stevemar: might as well move the doc change under that one so can see if it helps | 21:49 |
| morganfainberg | dhellmann, i dont want to halt work on kds while we split it out, because there is potentially TC things to be involved as well. | 21:50 |
| morganfainberg | dhellmann, so, lets see how things go. :) | 21:50 |
| morganfainberg | dhellmann, in either case thanks ! I'll def checkout the graduate script | 21:50 |
| morganfainberg | ayoung, ooh | 21:51 |
| *** pmathews has joined #openstack-dev | 21:51 | |
| ayoung | thought you would like that | 21:51 |
| morganfainberg | ayoung, that has potential... | 21:51 |
| morganfainberg | ayoung, esp. w/ the endpoint filter thing | 21:51 |
| ayoung | morganfainberg, it only made sense once we hade endpoint filtering, but now that we do.... | 21:51 |
| morganfainberg | when that goes in, i'd advocate graduating endpoint filtering out of being an extension | 21:51 |
| ayoung | morganfainberg, ++ but it can be done completely in keystoneclient | 21:52 |
| *** jhesketh_ has joined #openstack-dev | 21:52 | |
| *** jhesketh__ has joined #openstack-dev | 21:52 | |
| morganfainberg | ayoung, sure | 21:52 |
| *** bdpayne has quit IRC | 21:52 | |
| dolphm | morganfainberg: o/ we have a problem | 21:52 |
| morganfainberg | dolphm, ? | 21:52 |
| dolphm | morganfainberg: scroll to the very bottom http://logs.openstack.org/43/70843/1/gate/gate-tempest-dsvm-full/38732a1/logs/screen-key.txt.gz | 21:52 |
| Apsu | ahoy | 21:52 |
| dolphm | for reference- https://bugs.launchpad.net/devstack/+bug/1253482 | 21:53 |
| uvirtbot | Launchpad bug 1253482 in devstack "Keystone's IANA-assigned default port in linux local ephemeral port range" [Undecided,In progress] | 21:53 |
| morganfainberg | dolphm, 35357? again. | 21:53 |
| *** bdpayne has joined #openstack-dev | 21:53 | |
| morganfainberg | dolphm, yeah =/ | 21:53 |
| Apsu | morganfainberg: tl;dr, binding to 127.0.0.1 with all endpoints on 127.0.0.1 doesn't prevent port overlap, just reduces it | 21:53 |
| *** dvarga has quit IRC | 21:53 | |
| ayoung | ooops | 21:53 |
| morganfainberg | dolphm, Apsu, i thought that was the case. | 21:53 |
| Apsu | tl;tl;dr, local -> local connections with the client not specifying its source IP will pick the same IP as the destination. | 21:53 |
| Apsu | Meaning with an endpoint of 127.0.0.1, the client source will be 127.0.0.1 with a random ephemeral port | 21:54 |
| dolphm | Apsu: logstash queries appear to support the claim that the frequency of this bug has been reduced quite a bit | 21:54 |
| Apsu | dolphm: +1 | 21:54 |
| Apsu | Binding to *:$port means $port is unusable by anyone at any time on any IP | 21:54 |
| Apsu | Binding to 127.0.0.1:$port means 127.0.0.1:$port is unusable. | 21:54 |
| Apsu | If all connections are localhost -> localhost, the frequency would be the same. | 21:54 |
| bknudson | on my system devstack sets the system's IP address and not 127.0.0.1 | 21:54 |
| morganfainberg | is there anyway we can get the devstack gate vms to ... have the proper ephemeral (IANA spec) range? | 21:54 |
| Apsu | If some aren't, it'd be less. | 21:55 |
| Apsu | morganfainberg: Easier solution is to poke a hole with the ip_local_reserved_ports sysctl. | 21:55 |
| *** MaxV has quit IRC | 21:55 | |
| morganfainberg | Apsu, same net challenge | 21:55 |
| Apsu | sysctl net.ipv4.ip_local_reserved_ports=35357 | 21:55 |
| clarkb | Apsu: that isn't a solution | 21:55 |
| morganfainberg | Apsu, has to be done before devstack starts, part of the VM creation. | 21:55 |
| Apsu | morganfainberg: Is it? What is the challenge exactly? | 21:55 |
| Apsu | clarkb: Well it's *a* solution. Maybe not the best one :P | 21:55 |
| *** caleb_ has joined #openstack-dev | 21:55 | |
| clarkb | Apsu: it doesn't fix the problem there fore not a solution | 21:56 |
| clarkb | it mitigates it | 21:56 |
| dolphm | Apsu: how does it not fix it? | 21:56 |
| dolphm | clarkb: * | 21:56 |
| morganfainberg | it would reduce it for devstack to do that, but it's not the right fix for the devstack scripts to "fix" this | 21:56 |
| Apsu | Excuse my ignorance of the processes/infrastructure/politics here, I just understand the networking really well. | 21:56 |
| clarkb | dolphm: Apsu: because you need that in place before boot, devstack happens after boot | 21:56 |
| Apsu | You don't need it before boot. | 21:56 |
| bknudson | Apsu: do you know why linux has ephemeral port range outside of iana assigned? | 21:56 |
| dolphm | clarkb: right; is that not an option at all? | 21:56 |
| Apsu | It can change on the fly. | 21:56 |
| SergeyLukjanov | morganfainberg, I've posted some questions/suggestions to https://review.openstack.org/#/c/73074/ | 21:56 |
| *** ijw_ has quit IRC | 21:57 | |
| dolphm | Apsu: to be 100% effective, you do though - correct? | 21:57 |
| clarkb | Apsu: yes but it won't change existing connections aiui | 21:57 |
| clarkb | Apsu: so there is a race | 21:57 |
| Apsu | Unless you're saying "because after boot any process might be using the port" | 21:57 |
| Apsu | Fair enough. | 21:57 |
| clarkb | dolphm: no because it needs to happen in devstack | 21:57 |
| Apsu | It'd need to be in the VM image. | 21:57 |
| clarkb | dolphm: we want to test devstack not the CI vm build scripts | 21:57 |
| *** ijw has joined #openstack-dev | 21:57 | |
| dolphm | FWIW, the last 14 day query i ran for this bug 16 hits; i'm now getting 3 (and one of those occurrences was prior to the fix) | 21:57 |
| morganfainberg | if the VM is configured with the correct ephemeral ports (like keystone documents) then it is a solution. | 21:57 |
| *** rodrigods has quit IRC | 21:57 | |
| uvirtbot | Launchpad bug 16 in launchpad ""Swedish" and "Swedish (Sweden)" should be the same language" [Medium,Fix released] https://launchpad.net/bugs/16 | 21:57 |
| *** jecarey has quit IRC | 21:58 | |
| Apsu | morganfainberg: I see that as an identical challenge with more drawbacks than the reserved port approach. | 21:58 |
| morganfainberg | but i can see why that is a lot more overhead than having keystone do something different. | 21:58 |
| morganfainberg | e.g. httpd | 21:58 |
| Apsu | I.e., reducing the ephemeral range is always worse, due to restricting control plane headroom. | 21:58 |
| morganfainberg | but when linux and IANA are in conflict, there really is only so much we can do. | 21:59 |
| *** kevinconway has quit IRC | 21:59 | |
| Apsu | But poking a single hole has no practical impact. | 21:59 |
| dolphm | clarkb: i'd still argue that setting a ip_local_reserved_ports for 35357 would be a best practice for keystone deployments; so even if it happens in devstack and reduces the transient frequency further, that would be great | 21:59 |
| *** rmk has quit IRC | 21:59 | |
| clarkb | dolphm: I thought it was best to apache + wsgi? | 21:59 |
| dolphm | clarkb: ++ but you run into the same issue with apache starting up | 21:59 |
| Apsu | Other alternatives are much much uglier. By default on Linux, on all kernel versions, if you make a connection from a local process to another local process, on a local IP, the connecting process will choose the destination IP as its source IP, and an ephemeral port to bind to. | 21:59 |
| clarkb | dolphm: why? apache listens on 443 | 22:00 |
| Apsu | So no matter what IP you bind/connect to, you'll always be using the same one to connect from. | 22:00 |
| clarkb | and communication via wsgi isn't over a tcp socket is it? | 22:00 |
| dolphm | clarkb: fair enough! :) | 22:00 |
| Apsu | There are ways to 'hack' around that | 22:00 |
| *** mlegault has joined #openstack-dev | 22:00 | |
| Apsu | But I'm not sure you'll like any of them :P | 22:00 |
| *** yeylon__ has quit IRC | 22:00 | |
| clarkb | Apsu: right, the solution is to not use 35357 :) | 22:00 |
| dolphm | so i'm hearing that the next step is to stop running keystone outside of apache, period? | 22:00 |
| Apsu | That's one possibility. | 22:01 |
| dolphm | run apache by default in devstack, etc | 22:01 |
| clarkb | dolphm: that is what I was told by ayoung | 22:01 |
| Apsu | What image is the VM that's being used to test with? | 22:01 |
| dolphm | clarkb: i know he's an advocate, and i don't disagree | 22:01 |
| clarkb | dolphm: as apache is already the defacto default for keystone we just don't test it | 22:01 |
| ayoung | LIES! | 22:01 |
| clarkb | which is ugh | 22:01 |
| dolphm | ayoung: ? | 22:01 |
| Apsu | As in what distro/kernel version | 22:01 |
| ayoung | What | 22:01 |
| ayoung | Heh | 22:01 |
| *** pleia2 has quit IRC | 22:01 | |
| clarkb | Apsu: currently precise with latest 3.2 kernel | 22:01 |
| ayoung | I just heard my name | 22:01 |
| dolphm | ayoung: i'm taking words out of your mouth and passing them out | 22:01 |
| clarkb | will by trusty in when that happens | 22:01 |
| *** CrackerJackMack has quit IRC | 22:01 | |
| ayoung | Don't yuse 35357 | 22:01 |
| dolphm | ayoung: but correcting the spelling | 22:02 |
| dolphm | your welcome | 22:02 |
| dolphm | ;) | 22:02 |
| Apsu | clarkb: So... something fun you can do on newer kernels (meaning 3.x+) is replace the local route for the bind IP (ip r sh table local, to see them), exchanging the "src" hint with a different IP that's also on the box | 22:02 |
| *** ijw has quit IRC | 22:02 | |
| Apsu | And now when you make a local -> local connection to that target IP, it'll use the src hint IP as the source for the connection | 22:02 |
| ayoung | clarkb, yooose fower fower three | 22:03 |
| clarkb | Apsu: right but we don't want to do that | 22:03 |
| ayoung | dolphm, heh | 22:03 |
| ayoung | I need to finish token compression, I belive, in order to make Apache vialble long run | 22:03 |
| ayoung | guess what I am working on right now | 22:03 |
| Apsu | Bit hacky, but relatively easy and can happen after the boot | 22:03 |
| *** xingchao has quit IRC | 22:03 | |
| clarkb | Apsu: because silly work arounds are silly | 22:03 |
| clarkb | should be able to just run the service :) | 22:03 |
| *** ijw has joined #openstack-dev | 22:03 | |
| *** cburgess has quit IRC | 22:03 | |
| devoid | asking here because #openstack-nova never replied: do the libvirt driver maintainers meet at the general nova meting times? | 22:03 |
| dolphm | clarkb: implemented your suggestions on https://review.openstack.org/#/c/69997/2/queries/1253482.yaml to better track this bug | 22:03 |
| Apsu | clarkb: I figured. And yes. But if changing the VM is difficult... I find that a little silly too. Linux ephemerals may not be IANA, but Linux is super non-standard in hundreds of ways. | 22:03 |
| *** mlegault_ has joined #openstack-dev | 22:03 | |
| *** xingchao has joined #openstack-dev | 22:03 | |
| ayoung | clarkb, yeah, we should not have requeste a port in the ephemeral range....and IANAL should not be issuing them, either | 22:04 |
| Apsu | clarkb: So I find it a bit silly to insist on working around Linux ephemerals when Linux is, y'know, the de facto standard :P | 22:04 |
| dolphm | ayoung: good point | 22:04 |
| ayoung | ore Linux should change the range of hte ephemeral to match the standard | 22:04 |
| ayoung | or I should learn to type | 22:04 |
| Apsu | ayoung: That's very doubtful. | 22:04 |
| clarkb | dolphm: thanks, +2 | 22:04 |
| ayoung | dolphm, , https://blueprints.launchpad.net/python-keystoneclient/+spec/validate-service-endpoint that can happen at any point, since it is client only, right? | 22:04 |
| clarkb | Apsu: it isn't difficult to change the VM, it is the wrong thing to do | 22:05 |
| *** morganfainberg has quit IRC | 22:05 | |
| clarkb | Apsu: because $person is going to run devstack and if fails for them that is a fail | 22:05 |
| clarkb | same story for $person running a production keystone | 22:05 |
| *** CrackerJackMack has joined #openstack-dev | 22:05 | |
| Apsu | Sure. And that's fair. | 22:06 |
| *** vijendar has quit IRC | 22:06 | |
| Apsu | So, +1 for not using a port in the default Linux ephemeral. | 22:06 |
| *** cburgess has joined #openstack-dev | 22:06 | |
| *** pleia2 has joined #openstack-dev | 22:06 | |
| *** morganfainberg has joined #openstack-dev | 22:07 | |
| Apsu | Kind of why it exists, to avoid the local -> local port clobbering. | 22:07 |
| *** mlegault has quit IRC | 22:07 | |
| *** gokrokve has quit IRC | 22:07 | |
| *** ykhodork has quit IRC | 22:07 | |
| Apsu | Services are < ephem always :D | 22:07 |
| *** gokrokve has joined #openstack-dev | 22:08 | |
| *** rmk has joined #openstack-dev | 22:08 | |
| *** rmk has quit IRC | 22:08 | |
| *** rmk has joined #openstack-dev | 22:08 | |
| *** MaxV has joined #openstack-dev | 22:08 | |
| *** gokrokve_ has joined #openstack-dev | 22:08 | |
| *** rodrigods has joined #openstack-dev | 22:08 | |
| *** rodrigods has quit IRC | 22:08 | |
| dolphm | ayoung: yes, but i think that represents a misunderstanding of the catalog itself | 22:09 |
| *** doug_shelley66 has quit IRC | 22:09 | |
| *** xarses has quit IRC | 22:10 | |
| ayoung | dolphm, I think it more reflects a different use of it | 22:10 |
| ayoung | dolphm, with endpoint filtering, you can then say "only these endpoints can interoperate" | 22:11 |
| *** lucas-afk is now known as lucasagomes | 22:11 | |
| ayoung | dolphm, so you disagree? | 22:11 |
| dolphm | ayoung: ah, i forgot about endpoint filtering as this was the missing piece to make that even remotely useful | 22:11 |
| ayoung | ;0 | 22:11 |
| dolphm | ayoung: so yeah, that's viable then | 22:12 |
| *** jecarey has joined #openstack-dev | 22:12 | |
| * dolphm afk | 22:12 | |
| *** gokrokve has quit IRC | 22:12 | |
| *** mfer has quit IRC | 22:13 | |
| *** jamieh has quit IRC | 22:14 | |
| *** kbrierly has quit IRC | 22:14 | |
| morganfainberg | dolphm, https://review.openstack.org/#/c/73074/ | 22:14 |
| *** markmcclain has joined #openstack-dev | 22:14 | |
| *** ayoung is now known as ayoung-afk | 22:15 | |
| *** ayoung-afk is now known as ayoung-away | 22:15 | |
| *** nmagnezi has quit IRC | 22:21 | |
| *** ArxCruz has quit IRC | 22:24 | |
| *** xarses has joined #openstack-dev | 22:24 | |
| *** kbrierly has joined #openstack-dev | 22:25 | |
| *** zul has quit IRC | 22:26 | |
| *** tkay has quit IRC | 22:27 | |
| *** dsirrine is now known as dsirrine|afk | 22:29 | |
| *** athomas has quit IRC | 22:29 | |
| *** tkay has joined #openstack-dev | 22:29 | |
| *** tongli has quit IRC | 22:29 | |
| *** mdurnosvistov_ has quit IRC | 22:30 | |
| *** jnoller has quit IRC | 22:31 | |
| stevemar | bknudson, looks like docs is building now | 22:34 |
| bknudson | stevemar: how do you know? | 22:36 |
| *** jdob has quit IRC | 22:37 | |
| *** jay-lau-513 has quit IRC | 22:41 | |
| *** zzelle has quit IRC | 22:42 | |
| *** jmckind has quit IRC | 22:43 | |
| morganfainberg | bknudson, stevemar, dolphm, ayoung-away, jamielennox, lbragstad, dstanek, - OS Atlanta Summit, looks like I'm in the Omni (hotel). Not sure if you guys are looking at that yet or not, but - figured i'd let yall know | 22:43 |
| *** shardy is now known as shardy_afk | 22:43 | |
| lbragstad | morganfainberg: good to know, thanks! | 22:44 |
| morganfainberg | lbragstad, omni is also across the street from the conf basically | 22:44 |
| lbragstad | cool, so walking distance | 22:44 |
| *** e0ne has joined #openstack-dev | 22:46 | |
| *** e0ne has quit IRC | 22:48 | |
| *** sweston has quit IRC | 22:48 | |
| *** bhuvan has quit IRC | 22:49 | |
| *** mfer has joined #openstack-dev | 22:49 | |
| *** promulo has quit IRC | 22:50 | |
| morganfainberg | lbragstad, westin is a few bvlocks, so also walking distance | 22:50 |
| *** promulo has joined #openstack-dev | 22:50 | |
| morganfainberg | lbragstad, but the omni is def a lot closer | 22:50 |
| *** bhuvan has joined #openstack-dev | 22:50 | |
| * lbragstad takes notes | 22:51 | |
| *** Gordonz has quit IRC | 22:51 | |
| morganfainberg | lbragstad, westin price is $133/night, omni is $189/night - but afaict you can't go wrong with either | 22:51 |
| *** CaptTofu has quit IRC | 22:52 | |
| *** harlowja_away is now known as harlowja | 22:54 | |
| *** doug_shelley66 has joined #openstack-dev | 22:54 | |
| *** alop has quit IRC | 22:55 | |
| jamielennox | ayoung-away, morganfainberg: that blueprint you were tlking about: https://blueprints.launchpad.net/python-keystoneclient/+spec/validate-service-endpoint | 22:56 |
| jamielennox | do you know who Ian is (irc nick), because some of the changes i've got coming will make it more or less trivial | 22:57 |
| *** dims has quit IRC | 22:57 | |
| *** baoli has quit IRC | 22:58 | |
| *** novas0x2a|laptop has quit IRC | 22:59 | |
| *** mfer has quit IRC | 22:59 | |
| *** abhirc has joined #openstack-dev | 22:59 | |
| ayoung-away | jamielennox, yeah, he's in #moc | 23:00 |
| ayoung-away | jamielennox, yeha, it should be trivial | 23:00 |
| *** carl_baldwin has quit IRC | 23:01 | |
| jamielennox | moc? | 23:01 |
| *** stevemar has quit IRC | 23:01 | |
| *** abhirc has quit IRC | 23:02 | |
| *** abhirc has joined #openstack-dev | 23:02 | |
| *** SumitNaiksatam has quit IRC | 23:02 | |
| *** networks_ has quit IRC | 23:04 | |
| *** JoshuaG_AIM has quit IRC | 23:04 | |
| *** vartom1111111113 has quit IRC | 23:05 | |
| ayoung-away | Massachusetts Open Cloud. | 23:05 |
| *** xarses has quit IRC | 23:05 | |
| *** peristeri has quit IRC | 23:05 | |
| ayoung-away | jamielennox, but he's gone now | 23:05 |
| *** AlanClark has quit IRC | 23:05 | |
| jamielennox | ok, i'd like to have a chat with him about it at some point because session is changing how that works | 23:06 |
| *** rwsu has quit IRC | 23:06 | |
| morganfainberg | dolphm, https://launchpad.net/python-kiteclient ? | 23:07 |
| *** rwsu has joined #openstack-dev | 23:08 | |
| jamielennox | morganfainberg: i do like the name kite | 23:08 |
| morganfainberg | ayoung-away, ping when you have a sec | 23:08 |
| *** ayoung-away is now known as ayoung | 23:08 | |
| *** bhuvan has quit IRC | 23:08 | |
| ayoung | I'm here | 23:08 |
| morganfainberg | jamielennox, https://review.openstack.org/#/c/73022/1 https://review.openstack.org/#/c/73074/ | 23:09 |
| morganfainberg | ayoung, so. i am trying to resolve how to get to the point where revocation events and dogpile are in place | 23:09 |
| morganfainberg | ayoung, since i need both for ephemeral keys i think... | 23:09 |
| jamielennox | morganfainberg: kite-core? | 23:09 |
| ayoung | yep | 23:10 |
| morganfainberg | be back in a sec, need to take a call | 23:10 |
| morganfainberg | give me ~30min | 23:10 |
| dolphm | morganfainberg: will do | 23:10 |
| morganfainberg | can discuss that then | 23:10 |
| dolphm | jamielennox: probably | 23:10 |
| *** dims has joined #openstack-dev | 23:10 | |
| ayoung | jamielennox, I think that the endpoint enforcement is trivial. The endpoint needs to know its own Identifier | 23:10 |
| ayoung | is there any way to deduce that, or should it just go in the config file | 23:10 |
| jamielennox | ayoung: if we get session in then the only way to talk to an endpoint will be via the catalog so enforcement by default | 23:11 |
| *** marcoemorais has quit IRC | 23:11 | |
| *** marcoemorais has joined #openstack-dev | 23:11 | |
| *** yamahata has quit IRC | 23:11 | |
| *** alop has joined #openstack-dev | 23:11 | |
| ayoung | jamielennox, still want to enforce it on the server side (auth_token) | 23:12 |
| jamielennox | SUMMIT APPROVED!! aww yea | 23:12 |
| ayoung | Schweeet! | 23:12 |
| jamielennox | ayoung: is auth_token a problem though? we don't have auth_token talking to the identity endpoint from a tokne | 23:13 |
| jamielennox | from a service catalog | 23:13 |
| jamielennox | do we want it to? | 23:13 |
| jamielennox | that makes it difficult in terms of we have to fetch the right certs for that identity endpoint | 23:14 |
| dolphm | morganfainberg: done https://launchpad.net/python-kiteclient | 23:15 |
| *** yamahata has joined #openstack-dev | 23:15 | |
| ayoung | jamielennox, no, but say a user is doing direct curl, they send a token along, and auth token can enforce that the endpouint is specified in the service catalog of the token | 23:16 |
| *** flaper87 is now known as flaper87|afk | 23:16 | |
| *** otherwiseguy has quit IRC | 23:17 | |
| jamielennox | dolphm: i get to create a client from scratch? can't tell if i'm excited or the pressure is getting to me | 23:17 |
| *** xarses has joined #openstack-dev | 23:17 | |
| dolphm | jamielennox: lol i was just about to tell you that lol | 23:18 |
| jamielennox | ayoung: so just that the endpoint in the service catalog matches the one that auth_token is set up to use? | 23:18 |
| *** krtaylor has quit IRC | 23:19 | |
| *** avishay_away has quit IRC | 23:21 | |
| ayoung | jamielennox, each endpoint that enforces this would have a config vale "endpoint_id" and then the check is just that "endpoint_id" matches one of the endpoints in the service catalog of the token | 23:21 |
| *** avishay_away has joined #openstack-dev | 23:21 | |
| *** dstanek has quit IRC | 23:21 | |
| *** kgriffs has quit IRC | 23:21 | |
| ayoung | no cert fetch | 23:21 |
| *** tkay has quit IRC | 23:21 | |
| ayoung | although I do have a solution to figuring out which cert was used to sign a token, too | 23:21 |
| *** lbragstad has quit IRC | 23:22 | |
| jamielennox | ayoung: i'm interested in the second - but what's this talk of an endpoint_id | 23:22 |
| ayoung | jamielennox, right now an endpoint, say nova, does not know what its endpoint id is. | 23:23 |
| *** lucasagomes has quit IRC | 23:23 | |
| *** tkay has joined #openstack-dev | 23:23 | |
| *** mhu has quit IRC | 23:23 | |
| ayoung | So say I have a token with one nova endpoint in the service catalog, and I want to make sure it is only use on that endpoint, when I pass the token to any endpoint, auth_token on that endpoint needs to look at the service catalog and see "am I in there" | 23:24 |
| ayoung | It has to be enforced by auth_token | 23:24 |
| ayoung | its another form of token validity... | 23:24 |
| jamielennox | oh, ok | 23:24 |
| jamielennox | hmmm | 23:24 |
| *** armax has quit IRC | 23:25 | |
| *** markmcclain has quit IRC | 23:25 | |
| *** kgriffs has joined #openstack-dev | 23:25 | |
| ayoung | https://github.com/openstack/python-keystoneclient/blob/master/keystoneclient/middleware/auth_token.py#L842 jamielennox right after token bind...think "end point bind" | 23:25 |
| jamielennox | yea, that's what i was thinking | 23:26 |
| jamielennox | so it means having auth_token know its own endpoint_id | 23:26 |
| jamielennox | and it means having the endpoint_id exposed in the service catalog which we don't do at the moment | 23:27 |
| *** mhu has joined #openstack-dev | 23:27 | |
| jamielennox | the second one is more my concern there | 23:27 |
| *** mriedem has joined #openstack-dev | 23:27 | |
| dolphm | jamielennox: one auth_token configuration currently protects multiple endpoints | 23:27 |
| *** dstanek has joined #openstack-dev | 23:27 | |
| ayoung | ah...I thought it was | 23:27 |
| ayoung | it is in the smaple code in the identity-api | 23:27 |
| jamielennox | dolphm: protects? | 23:27 |
| dolphm | jamielennox: sits in front of? | 23:27 |
| jamielennox | oh, right public/internal urls are generally the same actual endpoint | 23:28 |
| dolphm | jamielennox: other services only have one auth_token filter configuration in their paste configs, and deploy that configured instance into multiple pipelines (public / admin / whatever) | 23:28 |
| *** hdd is now known as hdd_ | 23:28 | |
| dolphm | jamielennox: endpoint_id's should have *never* been exposed to the token | 23:28 |
| *** tkay has quit IRC | 23:28 | |
| ayoung | jamielennox, https://github.com/openstack/identity-api/blob/master/openstack-identity-api/v3/src/markdown/identity-api-v3.md#authentication-responses scroll down abit | 23:28 |
| *** byeager has quit IRC | 23:28 | |
| dolphm | that was just an oversight | 23:29 |
| ayoung | dolphm, but, but...they are useful | 23:29 |
| ayoung | can we keep them, please? | 23:29 |
| dolphm | ayoung: we don't have much choice, unless we're replacing the entire catalog... which we should also do | 23:29 |
| dolphm | but that's a bumpy road | 23:29 |
| ayoung | heh...so it is going to have to be "one of the endpoint ids" or are we ok with keeping it as one per? | 23:30 |
| *** Shaan7 has quit IRC | 23:30 | |
| dolphm | with minimal gains give ?nocatalog + GET /v3/catalog is a better approach | 23:30 |
| dolphm | ayoung: a list of endpoint id's gives the deployer flexibility | 23:30 |
| dolphm | ayoung: you could also just do it by actual endpoint value, which is what i would prefer | 23:31 |
| dolphm | (by URL) | 23:31 |
| *** thedodd has quit IRC | 23:31 | |
| ayoung | dolphm, can you add that to the BP, and your rationale? | 23:31 |
| dolphm | ayoung: link? | 23:31 |
| ayoung | dolphm, , https://blueprints.launchpad.net/python-keystoneclient/+spec/validate-service-endpoint | 23:33 |
| *** bauzas has quit IRC | 23:35 | |
| *** amerine_ is now known as amerine | 23:36 | |
| *** aeperezt has quit IRC | 23:36 | |
| *** aeperezt has joined #openstack-dev | 23:37 | |
| *** mlegault_ has quit IRC | 23:37 | |
| jamielennox | dolphm: i'm good with a new service catalog, the one we have is too big but i think that is best part of a new token format | 23:39 |
| jamielennox | ayoung: i don't see the need for exposing endpoint_id - other than the one just raised | 23:39 |
| *** achampion has quit IRC | 23:41 | |
| *** sarob has joined #openstack-dev | 23:41 | |
| *** yassine has quit IRC | 23:41 | |
| *** aeperezt has quit IRC | 23:42 | |
| *** devoid has quit IRC | 23:42 | |
| *** yamahata has quit IRC | 23:42 | |
| *** pixelb has quit IRC | 23:43 | |
| *** david-lyle has quit IRC | 23:43 | |
| *** derekh has quit IRC | 23:44 | |
| *** jmontemayor has quit IRC | 23:44 | |
| *** jmontemayor has joined #openstack-dev | 23:45 | |
| dolphm | ayoung: done | 23:45 |
| *** byeager has joined #openstack-dev | 23:46 | |
| *** jcoufal has quit IRC | 23:46 | |
| *** baoli has joined #openstack-dev | 23:46 | |
| *** byeager has quit IRC | 23:49 | |
| *** jmontemayor has quit IRC | 23:49 | |
| *** carl_baldwin has joined #openstack-dev | 23:50 | |
| *** baoli has quit IRC | 23:51 | |
| dolphm | ayoung: jamielennox: p.s. i'm not caught up on the mailing list thread yet, but i'm very much leaning towards the endpoint mutation hack, as long as it lands with clear documentation about how much of a hack it is lol | 23:51 |
| * ayoung hasn't read mail all day..kindof afraid to now | 23:51 | |
| *** cnesa has joined #openstack-dev | 23:52 | |
| ayoung | jamielennox, as far as the token<->cert thing | 23:52 |
| ayoung | it looks like poyasn is alrady in my venv. I tried using that tore read atoken, and I can find the signer info in it. | 23:52 |
| jamielennox | dolphm: mailing list thread? | 23:52 |
| ayoung | the Asn1 format is a little wonky | 23:52 |
| * jamielennox looking | 23:53 | |
| dolphm | jamielennox: ayoung's hack thread -- forgot the name | 23:53 |
| jamielennox | dolphm: oh that one | 23:53 |
| *** jckasper__ has joined #openstack-dev | 23:53 | |
| ayoung | ah...the whole "hack the version off the endpoint in the catalog" hack? | 23:53 |
| jamielennox | dolphm: i think the thread is going against you | 23:54 |
| dolphm | ayoung: yes | 23:54 |
| jamielennox | though i found something distressing yesterday | 23:54 |
| dolphm | jamielennox: probably -- i'm like a week behind on it (and all my email) | 23:54 |
| jamielennox | whilst GET / is unprotected by all servers, they pretty much unanimously need a token to access GET /vX | 23:54 |
| ayoung | oops | 23:55 |
| *** aeperezt has joined #openstack-dev | 23:55 | |
| ayoung | jamielennox, we don't | 23:55 |
| ayoung | are we unique in that? | 23:55 |
| jamielennox | ayoung: no and we can't | 23:55 |
| dolphm | jamielennox: p.s. i want better logging in the client :( | 23:55 |
| dolphm | jamielennox: wtf, really? | 23:55 |
| jamielennox | dolphm: patches welcome :) | 23:55 |
| dolphm | jamielennox: oh i guess so -- thanks auth_token! | 23:55 |
| *** CaptTofu has joined #openstack-dev | 23:56 | |
| * dolphm faceplam -- we should make an exception for that? | 23:56 | |
| jamielennox | dolphm: yea i get why it happened | 23:56 |
| jamielennox | but it's a problem | 23:56 |
| dolphm | definitely | 23:56 |
| jamielennox | it's workable if we start with the service catalog | 23:56 |
| jamielennox | because if you have a service catalog then you have a token and can just query with that | 23:57 |
| jamielennox | but it is definetly wrong and i'm not sure how it will affect us | 23:57 |
| *** jckasper_ has quit IRC | 23:57 | |
| dolphm | jamielennox: ++ | 23:57 |
| jamielennox | ayoung: do you want to have a pass at https://wiki.openstack.org/wiki/VersionDiscovery as well, i'll put it to -dev later today | 23:57 |
| *** gaelL has quit IRC | 23:58 | |
| dolphm | jamielennox: an authenticated request to GET /vX/ could return different results than an unauthenticated request, but you should be able to make both | 23:58 |
| * dolphm afk | 23:58 | |
| *** denis_makogon has quit IRC | 23:59 | |
| jamielennox | dolphm: right, that was the only thing i was thinknig that made it ok, if we are saying that the service catalog should contain unversioned endpoints maybe they want to customize the GET / requests per token instead | 23:59 |
| *** jckasper__ has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
