*** killer_prince has quit IRC | 00:22 | |
rloo | mrda: ha (or sob). I haven't yet gotten around to looking at the latest revision of the patch that 108974 is needed for. | 00:22 |
---|---|---|
mrda | yeah, me neither. Currently chasing red herrings all around the floor here. | 00:26 |
*** derekh_ has joined #openstack-ironic | 00:28 | |
rloo | mikal, 1347795 is a nova problem that is breaking things for tripleo and ironic. devananda said that this will fix (revert) so we're waiting... https://review.openstack.org/#/c/109033/ | 00:34 |
rloo | mikal: I see from 1347795 that you probably got the info you wanted. | 00:35 |
mikal | rloo: yeah, but the revert wasn't linked to the bug | 00:36 |
mikal | So I had trouble finding it | 00:36 |
mikal | I did get there in the end though | 00:36 |
rloo | mikal: not sure why it wasn't linked to the bug in the first place but glad you found it! | 00:37 |
mikal | rloo: the tag in the commit message is wrong is the problem | 00:37 |
rloo | mikal: yeah, just looked. Fixes-bug should be an alias to closes-bug! | 00:37 |
mikal | rloo: I dare you to propose a patch to jeepyb to do that... They'd take it I think. | 00:38 |
JayF | devananda: tl;dr update from the issue we were looking at: we removed about 10 bad BMCs from that group of servers, and the power status loop started working better. Still taking on the realm of 3-5 minutes to run. I think comstud has a patch coming that will add some parallellism to that. | 00:38 |
comstud | JayF, devananda: yes | 00:38 |
rloo | mikal: I learned many years ago, how to deal with dares. I ignore them. double-dares too ;) | 00:38 |
harlowja | rloo is very wise | 00:39 |
JayF | Now when I was a kid, I loved double-dare. Marc Summers was CRAZY! /s | 00:39 |
rloo | although I suppose someone could submit a bug/wishlist thingy. | 00:39 |
comstud | JayF, devananda: Essentially looking at something like this: http://paste.openstack.org/show/87861/ | 00:40 |
mikal | rloo: dammit | 00:42 |
mikal | rloo: management by goading is all I've got | 00:42 |
mikal | rloo: it would be impossible to handle to differnt but tags, totally impossible. | 00:43 |
rloo | mikal, how does mrda feel about that? :D | 00:43 |
mikal | rloo: ^-- let's see how that one works | 00:43 |
*** shakamunyi has joined #openstack-ironic | 00:43 | |
comstud | mikal: when you gonna start helping with OnMetal, mann? | 00:43 |
mikal | rloo: there is a high risk mrda will stab me in the neck with a pen sometime in the next week | 00:43 |
mikal | comstud: I am helping by stay a safe distance away | 00:43 |
comstud | :) | 00:43 |
jroll | lol | 00:44 |
* JayF adds mikal to the oncall rotation under 'third shift' | 00:44 | |
rloo | hmm, there might be more excitement at the meetup next week than I had thought. sorry I'm not going now ;) | 00:45 |
mikal | I am so busy that I am sure that if I signed up for anything I'd just be lying about my actual ability to contribute | 00:45 |
mikal | I'm therefore setting expectations low | 00:46 |
JayF | The best way you can help onmetal is by merging the Ironic -> Nova virt driver | 00:46 |
JayF | ;) | 00:46 |
mrda | mikal and rloo: there will be no stabbing, just rage quitting | 00:46 |
mikal | JayF: so, that would require mrda's patches to ironic to merge | 00:46 |
rloo | to be fair, mikal is helping. he's reviewed and provided comments. | 00:46 |
mikal | JayF: so you should review those... | 00:46 |
* mrda appreciates mikal's reviews | 00:46 | |
JayF | I'm just kidding, mostly :) | 00:46 |
mikal | I have also helped by not rage quitting | 00:47 |
mikal | Which has taken a lot of cups of tea | 00:47 |
jroll | "tea" | 00:47 |
mikal | Heh | 00:51 |
mikal | I dunno why, but I've been in a wine mood recently | 00:52 |
mrda | so tl;dr, everyone go review 107882 and 108872. Please :) | 00:52 |
mikal | Old age perhaps | 00:52 |
mrda | I think it's spent 'whine', mikal | 00:54 |
mrda | s/spent/spelt/ | 00:54 |
mrda | :-P | 00:54 |
rloo | s/spelt/spelled/ | 00:54 |
rloo | or did you mean spilt milk? | 00:54 |
mikal | Ugh | 01:00 |
mikal | You're driving me to Earl Gray here | 01:00 |
rloo | ok, I looked at 108872. Just needs another +2 on it. Not sure I'll get to 107882 tonight, it doesn't seem like bedtime reading to me. | 01:03 |
*** Haomeng|2 has joined #openstack-ironic | 01:04 | |
*** Haomeng has quit IRC | 01:05 | |
mrda | rloo: you are quite picky with your spelling, for an American. I see hardly any stray z's or missing u's :) | 01:10 |
rloo | mrda. Maybe that's cuz I'm not an American. | 01:11 |
mrda | that explains it! :) | 01:11 |
rloo | mrda. Are you saying Americans can't spell. I know they cannot pronounce the letter z and they have other quirks but... | 01:11 |
mrda | it's just the random substitution of z's for s's, and mostly the desire to drop u's out of words | 01:12 |
rloo | mrda. ha ha. yes, that's true. | 01:12 |
mrda | and I don't understand how 'caramel' can be pronounced as 'carmel' | 01:13 |
mikal | I fixed a bug in nova the other day that was caused by an extra u | 01:13 |
mikal | The queen's english strieks again | 01:13 |
mrda | strieks? Or shrieks? | 01:24 |
*** Haomeng|2 has quit IRC | 01:37 | |
*** nosnos has joined #openstack-ironic | 01:43 | |
rloo | mrda: I have a few minutes now. Are you OK with 107882? I don't want to look if you aren't. | 01:45 |
mrda | I always have time for you rloo :) | 01:46 |
mrda | I'm just lookin as to why the check-tempest-dvsm-virtual-ironic is afiling | 01:47 |
mrda | ^failing | 01:47 |
rloo | ha ha mrda. it looks like deva made changes to 107882. are you all ok with how the client is being imported etc? | 01:47 |
mrda | So yes, it appears to be working (but was so my __import__ techniaue too :) | 01:47 |
mrda | It's a good workaround until we can get those symbols exported from ironicclient | 01:48 |
rloo | mrda: so once those symbols are exported from ironicclient, then another change needs to be made? | 01:48 |
mrda | we have a python core contributor on my team, who unfortunately is at EuroPython right now. but once he reappears I'll work out how to fix ironicclient | 01:48 |
mrda | rloo: yes, we can unwind the hasattr | 01:49 |
mrda | or maybe not, because otherwise old ironicclients won't work | 01:49 |
mrda | so it'll have to stay for a while | 01:49 |
rloo | mrda: unwind in both nova/ironic trees? | 01:49 |
mrda | I still would like to fix ironicclient though | 01:49 |
*** jcoufal has joined #openstack-ironic | 01:49 | |
mrda | rloo: well, I think we leave it as is. There's no harm in deva's solution remaining there as is | 01:50 |
rloo | mrda: just wondering if we can fix ironicclient soon, and have the 'right' solution in the driver soon. or if we need to go with the temporary solution in the driver now. | 01:50 |
mrda | so the summary is that we can go ahead with it as is | 01:50 |
rloo | mrda. ok, i'll look at the review then ;) | 01:51 |
mrda | I think we leave it there and just go ahead. Old ironicclient's will still need support, for a hile, so it has to stay as is | 01:51 |
mrda | thanks rloo - appreciated! | 01:51 |
rloo | err, i guess i'll review the patch then. | 01:51 |
mrda | in the mean time, I'll look at -dvsm-virtual-ironic | 01:51 |
rloo | mrda: those tests are failing due to nova. unless they're failing for another reason too. | 01:52 |
rloo | mrda: waiting for https://review.openstack.org/#/c/109033/ to land. | 01:52 |
mrda | oh, I thought that had merged already | 01:52 |
* mrda looks | 01:52 | |
rloo | mrda. you can take a break and have some tea then :-) | 01:52 |
mrda | lol, coffee for me :) | 01:53 |
mrda | And there's still other nova ironic stuff to do anyways | 01:53 |
mrda | thanks rloo for your help! | 01:54 |
rloo | mrda, glad to know you have stuff to do ;) | 01:54 |
rloo | yw mrda | 01:54 |
* mrda notes that 109033 won't get through gate for another 30 mins at least | 01:55 | |
*** mdorman has quit IRC | 02:29 | |
*** killer_prince has joined #openstack-ironic | 02:33 | |
*** killer_prince is now known as lazy_prince | 02:33 | |
*** rloo has quit IRC | 02:43 | |
*** pcrews has quit IRC | 02:45 | |
*** vinbs has joined #openstack-ironic | 02:59 | |
*** ramineni has joined #openstack-ironic | 03:02 | |
*** derekh_ has quit IRC | 03:09 | |
*** nosnos has quit IRC | 03:15 | |
*** viktors|afk has quit IRC | 03:54 | |
*** krtaylor has joined #openstack-ironic | 04:17 | |
*** nosnos has joined #openstack-ironic | 04:18 | |
*** lazy_prince is now known as killer_prince | 04:21 | |
*** Poornima_ has joined #openstack-ironic | 04:26 | |
*** eghobo has joined #openstack-ironic | 04:36 | |
mikal | https://review.openstack.org/#/c/109033 failed to merge | 05:03 |
openstackgerrit | Michael Davies proposed a change to openstack/ironic: Import fixes from the Nova driver reviews https://review.openstack.org/107882 | 05:03 |
mrda | mikal: I cried when I saw that | 05:05 |
*** harlowja has quit IRC | 05:11 | |
*** harlowja has joined #openstack-ironic | 05:11 | |
*** eghobo has quit IRC | 05:19 | |
*** k4n0 has joined #openstack-ironic | 05:24 | |
*** bvivek has joined #openstack-ironic | 05:27 | |
*** eghobo has joined #openstack-ironic | 05:29 | |
openstackgerrit | Michael Davies proposed a change to openstack/ironic: Backporting nova host manager changes into ironic https://review.openstack.org/108872 | 05:35 |
mrda | devananda: Just on 103165 - I included Mikyung and Kurt because of their contributions to baremetal_host_manager.py ion the nova tree, which is a predecessor of the {base_baremetal, baremetal, and ironic}_host_manager.py in our tree | 05:38 |
*** rameshg87 has joined #openstack-ironic | 05:39 | |
*** rakesh_hs has joined #openstack-ironic | 05:43 | |
*** dkehn_ has joined #openstack-ironic | 05:45 | |
*** dkehn__ has joined #openstack-ironic | 05:47 | |
*** bmahalakshmi has joined #openstack-ironic | 05:47 | |
*** dkehnx has quit IRC | 05:49 | |
*** dkehn_ has quit IRC | 05:51 | |
*** marios has quit IRC | 05:59 | |
*** marios has joined #openstack-ironic | 06:02 | |
*** Halacs has joined #openstack-ironic | 06:08 | |
*** eghobo has quit IRC | 06:10 | |
openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/ironic: Imported Translations from Transifex https://review.openstack.org/108901 | 06:10 |
*** zigo has quit IRC | 06:27 | |
rameshg87 | JoshNang: hi | 06:29 |
JoshNang | rameshg87: hey, what's up? | 06:29 |
rameshg87 | JoshNang: :-) | 06:30 |
*** mitz has quit IRC | 06:30 | |
rameshg87 | JoshNang: going good. how are you ? | 06:30 |
rameshg87 | JoshNang: had a question regarding the devstack script in ironic which creates the virsh baremetal vms for testing. | 06:31 |
rameshg87 | JoshNang: do you use this ? | 06:31 |
JoshNang | pretty good! | 06:31 |
JoshNang | i haven't really. i'm planning to mess with it in the next couple days | 06:31 |
*** mitz has joined #openstack-ironic | 06:32 | |
rameshg87 | JoshNang: okay. | 06:32 |
rameshg87 | JoshNang: i was just trying out this: http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#deploying-ironic-with-devstack | 06:32 |
*** zigo has joined #openstack-ironic | 06:32 | |
rameshg87 | JoshNang: ends up in some issues for me. i had some doubts on how this works. | 06:32 |
*** ChanServ changes topic to "Bare Metal Provisioning | Status: http://bit.ly/ironic-whiteboard | Docs: http://docs.openstack.org/developer/ironic/ | Bugs: https://bugs.launchpad.net/ironic" | 06:33 | |
rameshg87 | JoshNang: looks like lucasagomes wrote these. i think i can check with him then.. | 06:33 |
JoshNang | rameshg87: what issues are you having? | 06:34 |
rameshg87 | JoshNang: i tried to deploy a vm with pxe_ssh driver (as mentioned in the guide above) | 06:35 |
rameshg87 | JoshNang: the virsh vms are connected to a ovs bridge brbm | 06:36 |
rameshg87 | JoshNang: while n-dhcp of neutron is listening on br-int | 06:36 |
rameshg87 | JoshNang: this says brbm and br-int are connected - https://github.com/openstack-dev/devstack/blob/master/lib/ironic#L382 | 06:36 |
*** Mikhail_D_ltp has joined #openstack-ironic | 06:36 | |
rameshg87 | JoshNang: but unfortunately the pxe boot request from virsh vms doesn't go to n-dhcp, and pxe boot fails | 06:37 |
*** killer_prince is now known as lazy_prince | 06:37 | |
JoshNang | hmmmm. i can't say i've ever had that problem unfortuantely | 06:37 |
rameshg87 | JoshNang: hmm..okay | 06:38 |
rameshg87 | JoshNang: let me try hard if i can see something else .. | 06:38 |
*** viktors has joined #openstack-ironic | 06:40 | |
*** harlowja is now known as harlowja_away | 06:42 | |
JoshNang | rameshg87: i'm gonna head to bed...early morning tomorrow. i'll be testing some stuff in devstack in the next couple days, so if you don't find a solution soon, maybe i'll be able to help more | 06:47 |
rameshg87 | JoshNang: thanks :-) | 06:47 |
rameshg87 | JoshNang: will continue to look more and ping you if i am still stuck up with issues | 06:48 |
JoshNang | kk! good luck! | 06:48 |
*** shakamunyi has quit IRC | 07:03 | |
*** jcoufal has quit IRC | 07:16 | |
*** bvivek has quit IRC | 07:23 | |
*** jistr has joined #openstack-ironic | 07:24 | |
*** yuriyz has quit IRC | 07:25 | |
*** krtaylor has quit IRC | 07:25 | |
mrda | Night all | 08:05 |
*** mrda is now known as mrda-away | 08:05 | |
*** romcheg1 has joined #openstack-ironic | 08:09 | |
*** yuriyz has joined #openstack-ironic | 08:11 | |
*** kpavel has joined #openstack-ironic | 08:15 | |
*** pelix has joined #openstack-ironic | 08:23 | |
*** lucasagomes has joined #openstack-ironic | 08:23 | |
dtantsur|afk | Morning, Ironic! | 08:23 |
*** dtantsur|afk is now known as dtantsur | 08:23 | |
*** bvivek has joined #openstack-ironic | 08:25 | |
jroll | I know I'm up too late when I can say good morning to dtantsur :p | 08:50 |
dtantsur | lol | 08:50 |
dtantsur | g'evening, jroll, what's the time for you? | 08:50 |
romcheg1 | dtantsur: I think it's almost morning for jroll too :) | 08:51 |
jroll | approaching 2am, ha | 08:51 |
romcheg1 | Morning guys btw :) | 08:51 |
jroll | heya romcheg1 | 08:51 |
dtantsur | romcheg1, morning :) | 08:52 |
*** martyntaylor has joined #openstack-ironic | 09:20 | |
*** kpavel has quit IRC | 09:38 | |
*** Poornima_ has quit IRC | 09:46 | |
*** foexle has joined #openstack-ironic | 09:46 | |
*** datajerk has quit IRC | 09:49 | |
*** datajerk has joined #openstack-ironic | 09:49 | |
*** k4n0 has quit IRC | 10:07 | |
*** romcheg1 has quit IRC | 10:08 | |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Expose {set,get}_boot_device in the API https://review.openstack.org/90151 | 10:12 |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: SSH virsh to use the new ManagementInterface https://review.openstack.org/89884 | 10:12 |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: IPMINative to use the new ManagementInterface https://review.openstack.org/86588 | 10:12 |
*** k4n0 has joined #openstack-ironic | 10:21 | |
*** jbjohnso has quit IRC | 10:26 | |
*** jbjohnso has joined #openstack-ironic | 10:26 | |
*** krtaylor has joined #openstack-ironic | 10:33 | |
*** jbjohnso has quit IRC | 10:47 | |
*** jbjohnso has joined #openstack-ironic | 10:48 | |
*** martyntaylor has left #openstack-ironic | 10:53 | |
*** vinbs has quit IRC | 10:56 | |
*** Alexei_9871 has joined #openstack-ironic | 10:58 | |
*** harlowja_away has quit IRC | 10:59 | |
*** Poornima_ has joined #openstack-ironic | 10:59 | |
*** k4n0 has quit IRC | 11:01 | |
*** ramineni has quit IRC | 11:02 | |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Clean up calls to get_chassis() https://review.openstack.org/108959 | 11:05 |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Clean up calls to get_port() https://review.openstack.org/109252 | 11:05 |
rameshg87 | dtantsur: hi | 11:08 |
dtantsur | rameshg87, hi | 11:08 |
*** vdrok has joined #openstack-ironic | 11:08 | |
rameshg87 | dtantsur: i just encountered something which is thrown as a warning, but i feel it should be an error and should be caught in validate | 11:09 |
rameshg87 | dtantsur: https://github.com/openstack/ironic/blob/master/ironic/common/neutron.py#L148-L153 | 11:09 |
rameshg87 | dtantsur: the code that updates neutron with the dhcp options when provisioning with pxe | 11:09 |
rameshg87 | dtantsur: shouldn't it be error, because unless we update neutron, the pxe boot is not going to work, right ? | 11:09 |
dtantsur | rameshg87, not sure, but likely yes | 11:10 |
*** jcoufal has joined #openstack-ironic | 11:11 | |
rameshg87 | dtantsur: i feel we should validate atleast one of the ironic ports have a vif_port_id set | 11:11 |
*** Poornima_ has quit IRC | 11:14 | |
dtantsur | brb | 11:16 |
*** vdrok has quit IRC | 11:18 | |
*** vdrok_ has joined #openstack-ironic | 11:18 | |
*** k4n0 has joined #openstack-ironic | 11:19 | |
openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Implements send-data-to-ceilometer https://review.openstack.org/72538 | 11:23 |
*** bvivek has quit IRC | 11:30 | |
*** marzif has joined #openstack-ironic | 11:32 | |
*** Poornima_ has joined #openstack-ironic | 11:40 | |
*** lucasagomes is now known as lucas-hungry | 11:42 | |
openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Implements send-data-to-ceilometer https://review.openstack.org/72538 | 11:50 |
*** bmahalakshmi has quit IRC | 11:54 | |
openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Implements send-data-to-ceilometer https://review.openstack.org/72538 | 11:55 |
openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Implements send-data-to-ceilometer https://review.openstack.org/72538 | 11:56 |
*** k4n0 has quit IRC | 11:59 | |
*** cdent has joined #openstack-ironic | 12:00 | |
*** chuckC has quit IRC | 12:07 | |
*** nosnos has quit IRC | 12:09 | |
*** jcoufal has quit IRC | 12:19 | |
*** cdent_ has joined #openstack-ironic | 12:21 | |
*** rameshg87 has left #openstack-ironic | 12:22 | |
*** cdent has quit IRC | 12:22 | |
*** cdent_ is now known as cdent | 12:22 | |
marios | dtantsur: hey bud | 12:35 |
dtantsur | marios, hi | 12:35 |
marios | dtantsur: so following http://docs.openstack.org/developer/ironic/dev/dev-quickstart.html#deploying-ironic-with-devstack | 12:35 |
marios | devtest completes fine, everything running ok | 12:35 |
marios | can't deploy instance (as per the instructions there) with baremetal flavor | 12:36 |
marios | immediately goes to ERROR state | 12:36 |
dtantsur | marios, what ironic node-show shows about node? what nova show shows about instance? | 12:36 |
marios | dtantsur: ironic nodes are all powered-off | 12:37 |
marios | dtantsur: just wondering if you had seen this before mate (i am using one of our beaker boxes in brno for this) | 12:37 |
dtantsur | marios, I know quite a few reasons for this behavior, so trying to figure out which one you have :) | 12:38 |
dtantsur | marios, or you can give me access to the box and I'll investigate myself | 12:39 |
marios | dtantsur: i had a quick look through the screen sessions for n-cpu/i-api etc but couldn't see something obvious there | 12:39 |
dtantsur | marios, ir-cond logs are much more important | 12:40 |
dtantsur | well, may be | 12:40 |
*** ChuckC has joined #openstack-ironic | 12:40 | |
marios | dtantsur: ah didn't check that, going to try it again and watch the log | 12:40 |
dtantsur | marios, also before trying to deploy anything, you may try ironic set-node-power-state on, that's a bit easier to check | 12:41 |
*** lucas-hungry is now known as lucasagomes | 12:41 | |
dtantsur | you'll at least know, whether the driver can manager power | 12:41 |
lucasagomes | yeah ir-cond logs | 12:41 |
lucasagomes | also our gate is broken due a change in nova | 12:41 |
dtantsur | marios, even before, try ironic node-validate on a node | 12:41 |
marios | dtantsur: great thanks will try this all | 12:42 |
lucasagomes | that may be the cause of the error as well (the gate does pretty much the same steps there ^) | 12:42 |
*** Poornima_ has quit IRC | 12:42 | |
dtantsur | lucasagomes, oh, makes a lot of sense | 12:43 |
lucasagomes | yeah, it might worth try to reset nova | 12:43 |
lucasagomes | to the commit before that change (will grab the link) | 12:43 |
lucasagomes | and then run it | 12:43 |
lucasagomes | https://review.openstack.org/#/c/71557/33 that's the commit that broke us | 12:44 |
lucasagomes | marios, ^ | 12:44 |
marios | lucasagomes: thx bud | 12:44 |
lucasagomes | no worries | 12:44 |
*** jdob has joined #openstack-ironic | 12:48 | |
*** rloo has joined #openstack-ironic | 13:00 | |
*** zigo has quit IRC | 13:03 | |
*** jcoufal has joined #openstack-ironic | 13:03 | |
*** jcoufal has quit IRC | 13:04 | |
*** zigo has joined #openstack-ironic | 13:04 | |
*** jcoufal has joined #openstack-ironic | 13:07 | |
*** athomas has joined #openstack-ironic | 13:07 | |
*** _dguerri is now known as dguerri | 13:08 | |
openstackgerrit | Haomeng,Wang proposed a change to openstack/ironic: Implements send-data-to-ceilometer https://review.openstack.org/72538 | 13:10 |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Fix nodes left in an incosistent state if no workers https://review.openstack.org/100958 | 13:10 |
lucasagomes | Shrews, ^ had totally forgot to test it before... I left a comment with my findings inline | 13:10 |
jroll | hey y'all... today is spec freeze day. gotta land https://review.openstack.org/#/c/98506/ | 13:11 |
Shrews | lucasagomes: ah, i'd forgotten about that :) | 13:11 |
lucasagomes | Shrews, yeah me too, I just saw it and decided to give it a go here locally | 13:12 |
lucasagomes | it works fine | 13:12 |
lucasagomes | if u wanna try just register some machines and set the workers_pool_size=1 in the ironic.conf | 13:13 |
lucasagomes | then try to boot machine | 13:13 |
lucasagomes | s/boot/deploy | 13:13 |
*** matty_dubs|gone is now known as matty_dubs | 13:13 | |
lucasagomes | Shrews, and cleaning the instance_uuid won't free the resources in nova | 13:14 |
rloo | jroll: I thought today was the deadline for any new specs. hmm, let me check. | 13:14 |
lucasagomes | Shrews, we need to deleted the failed instance in nova in order to free those | 13:14 |
jroll | rloo: oh, maybe, idk | 13:14 |
lucasagomes | rloo, yeah it's for new ones | 13:14 |
jroll | rloo: we want to land the code next week, so in any case that spec needs to get through | 13:14 |
lucasagomes | but we gotta merge that anyway :) | 13:14 |
rloo | jroll: https://etherpad.openstack.org/p/3sxKH2po1o. "new spec proposal freeze" | 13:15 |
jroll | :) | 13:15 |
marios | lucasagomes: so, reverting nova has helped (nova list says 'spawning now', before was immediately error). | 13:15 |
lucasagomes | marios, ack... but it's stuck on the spawning ? | 13:15 |
lucasagomes | or did you go through it? | 13:15 |
rloo | jroll: yeah, that and the swift url thing is on my list to look at today. although the quality of my reviews seems to be deteriorating. I think I need a break... | 13:15 |
marios | ACTIVE \o/ thanks man | 13:15 |
lucasagomes | marios, :) no bothers | 13:15 |
lucasagomes | yw | 13:15 |
jroll | rloo: thanks :) | 13:15 |
lucasagomes | marios, that change in nova is being reverted because it also break nova bm | 13:16 |
lucasagomes | and consequently tripleO | 13:16 |
marios | ic | 13:16 |
devananda | morning, all | 13:16 |
lucasagomes | devananda, morning | 13:17 |
rloo | morning devananda. | 13:17 |
*** dkehn__ is now known as dkehnx | 13:17 | |
*** athomas has quit IRC | 13:21 | |
NobodyCam | good morning Ironic | 13:24 |
lucasagomes | morning NobodyCam | 13:25 |
NobodyCam | :) moring lucasagomes | 13:25 |
dtantsur | NobodyCam, morning | 13:25 |
jroll | mornin devananda | 13:25 |
NobodyCam | just walking up slowly and John out the door, on his way to the airport :-p | 13:25 |
NobodyCam | mornig dtantsur and jroll too :) | 13:26 |
dtantsur | devananda, jroll, rloo, morning as well | 13:26 |
rloo | morning dtantsur and all Ironickers out there :) | 13:27 |
NobodyCam | ya morning rloo :) | 13:27 |
rloo | including NobodyCam! | 13:27 |
NobodyCam | hheeheh :) | 13:27 |
devananda | NobodyCam: g'morning! | 13:28 |
devananda | NobodyCam: also, wow, I forget how long it takes to get anywhere from the peninsula :( | 13:28 |
NobodyCam | lol :) | 13:29 |
NobodyCam | ya | 13:29 |
jroll | which peninsula? | 13:30 |
NobodyCam | devananda: are you on a red eye tonight? | 13:30 |
devananda | NobodyCam: no. I fly out of Raleigh in ~5hr, get in around 9pm | 13:30 |
NobodyCam | :) | 13:30 |
NobodyCam | Johns the Hilton? | 13:31 |
NobodyCam | gah | 13:31 |
NobodyCam | John at the hilton, | 13:32 |
NobodyCam | he like me to ask what time it all starts tomorrow? | 13:32 |
devananda | John's THE hilton sounds so much better :p | 13:32 |
NobodyCam | lol | 13:32 |
devananda | NobodyCam: 8am | 13:32 |
NobodyCam | ack so will make sure his alarms are set befor leaving | 13:33 |
NobodyCam | :) | 13:35 |
*** vdrok_ has quit IRC | 13:37 | |
*** cflaskerud has joined #openstack-ironic | 13:41 | |
NobodyCam | jroll: olympic peninsula | 13:41 |
rloo | JoshNang: if you get this, could you please set the url for the spec for your bp: https://blueprints.launchpad.net/ironic/+spec/swift-temp-urls | 13:48 |
JoshNang | rloo: np! | 13:50 |
rloo | thx JoshNang. (what are you doing up so early.) | 13:50 |
*** eghobo has joined #openstack-ironic | 13:50 | |
Shrews | rloo: was there a bug for the nova revert that we could use for recheck? | 13:54 |
rloo | Shrews: yup. sec | 13:54 |
rloo | https://bugs.launchpad.net/ironic/+bug/1347795 | 13:54 |
rloo | ^^ Shrews. | 13:54 |
Shrews | rloo: gracias | 13:54 |
rloo | Shrews: but the patch didn't merge yet. | 13:55 |
Shrews | oh poo | 13:55 |
rloo | Shrews: so I don't see any reason to do a recheck. it'll still fail. | 13:55 |
JoshNang | rloo: updated with the spec url and fixed the description to be more relevant | 13:55 |
JoshNang | (getting an early start this morning) | 13:55 |
Shrews | rloo: ok. i was in and out yesterday dealing with a home situation, so i'm not fully up to speed on it. thx | 13:55 |
rloo | Shrews: looks like the patch needs to be rebased. So another day lost. And in the bug, someone says that the undo will fix the ci but there's a problem with the code still. | 13:56 |
rloo | Shrews. I've tried to keep our IronicWhiteBoard up-to-date. Sort of ;) | 13:56 |
Shrews | ah, will read | 13:56 |
devananda | so don't bother rechecking yet | 13:57 |
devananda | the fix hasn't landed in nova yet and our tests are guaranteed to fail until it does | 13:58 |
Shrews | passing dicts of info around is destined for failure. that should be a proper object | 13:58 |
rloo | devananda, Shrews: I just updated the etherpad to say don't bother to recheck ;) | 13:58 |
Shrews | object == contract == less breakage | 13:59 |
rloo | thx JoshNang! Gonna review your patch soon... | 14:01 |
*** cflaskerud has left #openstack-ironic | 14:02 | |
devananda | Shrews: ++ | 14:02 |
*** pcrews has joined #openstack-ironic | 14:03 | |
*** eghobo has quit IRC | 14:08 | |
openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: EXPERIMENTAL Add Conductor.discovery_driver field https://review.openstack.org/109304 | 14:14 |
JoshNang | rloo: thanks! :) | 14:17 |
*** rakesh_hs has quit IRC | 14:22 | |
*** jgrimm has joined #openstack-ironic | 14:26 | |
*** rwsu has joined #openstack-ironic | 14:28 | |
*** derekh_ has joined #openstack-ironic | 14:29 | |
*** shakamunyi has joined #openstack-ironic | 14:31 | |
Shrews | Can anyone explain the pep8 logic that says this import ordering makes sense? https://review.openstack.org/#/c/108872/4/ironic/nova/tests/scheduler/test_ironic_host_manager.py | 14:43 |
devananda | Shrews: yes. if you s/ironic.nova/nova/ then it makes sense | 14:44 |
devananda | Shrews: we aren't running pep8 on the nova code in our tree | 14:44 |
Shrews | devananda: ok, makes sense with the substitution | 14:45 |
Shrews | and since you just +A'd it, i guess i no longer need to keep reviewing it :) | 14:45 |
devananda | so I'm +A'ing several things, knowing that they won't pass right now | 14:46 |
NobodyCam | ack | 14:46 |
devananda | not until 109033 merges in Nova | 14:46 |
devananda | then we can reverify bug 1347795 on lots of things | 14:47 |
NobodyCam | +++ | 14:47 |
devananda | I'll be on a plane for ~10 hours, so want to get things in a state ya'll can finish merging them | 14:47 |
Shrews | "ya'll"? devananda, you've been in NC too long. LOL | 14:47 |
devananda | lol | 14:47 |
NobodyCam | lol | 14:48 |
jroll | nah, he would know it's spelled "y'all" if he had been there too long :P | 14:48 |
Shrews | if he knew what "you'uns" was, i'd be worried | 14:48 |
*** foexle has quit IRC | 14:48 | |
devananda | Shrews: that takes entirely too many letters | 14:48 |
jroll | JayF taught me about "bo-fum" yesterday | 14:49 |
JayF | yep because when you're fixing two things | 14:50 |
JayF | you gotta write configs to bo-fum | 14:50 |
*** romcheg1 has joined #openstack-ironic | 14:52 | |
*** shakamunyi has quit IRC | 14:53 | |
Shrews | JayF: that's again as much work to change bo-fum | 14:53 |
openstackgerrit | Dmitry Tantsur proposed a change to openstack/ironic: EXPERIMENTAL Implement conductor part of hardware discovery https://review.openstack.org/109312 | 14:53 |
*** Mikhail_D_ltp has quit IRC | 14:54 | |
dtantsur | folks, what are you talking about? :) | 14:54 |
Shrews | lol | 14:55 |
*** dguerri is now known as _dguerri | 15:02 | |
*** MattMan has quit IRC | 15:03 | |
* rloo is happy to be an ironicker. Gets to learn stuff like bofum. | 15:04 | |
*** MattMan has joined #openstack-ironic | 15:04 | |
NobodyCam | :) | 15:04 |
*** _dguerri is now known as dguerri | 15:07 | |
*** lazy_prince is now known as killer_prince | 15:07 | |
devananda | rloo: testing your client change against an api server w/o the new api | 15:17 |
devananda | rloo: i get this: Not Found (HTTP 404) | 15:17 |
devananda | which isn't terrible, but wondering if we should do something a little mroe helpful for the client | 15:17 |
devananda | so, regarding our J2 milestone | 15:23 |
devananda | I'm holding off on tagging it until either: 109033 lands to fix our gate, and we land a bunch of outstanding fixes. OR Friday afternoon, which ever comes first. | 15:23 |
jroll | onmetal just went live :) | 15:24 |
devananda | jroll: \o/ | 15:24 |
jroll | y'all should try ironic in production | 15:24 |
JayF | http://www.rackspace.com/blog/onmetal-shows-phenomenal-performance-in-early-benchmarks/ && http://www.marketwired.com/press-release/rackspace-announces-onmetal-cloud-servers-general-availability-and-pricing-nyse-rax-1932532.htm | 15:24 |
*** mdorman has joined #openstack-ironic | 15:24 | |
comstud | devananda: your thing is a thing | 15:24 |
comstud | not that it wasn't a thing before | 15:24 |
jroll | devananda: right back at ya ^ | 15:24 |
*** jistr has quit IRC | 15:24 | |
JayF | Now for the philosophical question: is true graduation productization? | 15:25 |
JayF | :P | 15:25 |
jroll | lol | 15:25 |
lucasagomes | o/ nice one | 15:25 |
*** pradipta_away has quit IRC | 15:25 | |
devananda | comstud: that's fantastic. I love that you have made my thing a thing. | 15:25 |
comstud | :) | 15:25 |
devananda | JayF: no. but it further lends credence to the argument I've been making that graduation has NOTHING to do with quality of code or who in the community is (or should be) using what | 15:25 |
devananda | integration is part of the development workflow, NOT a product statement | 15:26 |
devananda | though, today, it's often seen that way externally | 15:26 |
devananda | I want to decouple that | 15:26 |
NobodyCam | jroll: awesome-ness \o/ | 15:27 |
rloo | JayF, comstud, jroll, JoshNang, etc: congrats! | 15:27 |
JoshNang | rloo: thanks! | 15:27 |
comstud | ty | 15:27 |
JoshNang | rloo: also, working on your comments on swift right now | 15:27 |
jroll | thanks :) | 15:27 |
jroll | look at that | 15:27 |
comstud | now.. we do have some performance things to work on :) | 15:27 |
jroll | celebrating with an ironic patch | 15:27 |
lucasagomes | :D | 15:28 |
rloo | devananda: wrt the client change. what do you suggest? it'll be a similar problem with any new feature. Should that be a separate patch (or yikes a BP proposal)? | 15:28 |
devananda | comstud: definitely | 15:28 |
comstud | there's a big one | 15:28 |
comstud | that is in nova | 15:28 |
devananda | rloo: I'm not sure. perhaps 404 is totally fine. | 15:28 |
rloo | JoshNang: oh, I had another comment. I'll just ask you. Why do you use the swift_endpoint_url, instead of the one from the direct_url? | 15:28 |
devananda | rloo: and certainly i'm not suggesting any more BPs for this | 15:28 |
devananda | rloo: just wondering what the best clietn interaction is goign to be | 15:29 |
rloo | devananda: yahoo. no more bps for it. The client can't know whether that feature is available or not. if the client is *sure* that the url it generated is correct, then I suppose it can replace the msg with something else. I could spend a bit of time looking at what other clients do. | 15:30 |
Shrews | devananda: rloo: 404 seems reasonable to me | 15:36 |
rloo | Shrews: thx. Vote as you see fit. https://review.openstack.org/#/c/76338/ ;) | 15:38 |
*** rloo is now known as rloo_afk | 15:39 | |
rloo_afk | bbiab | 15:39 |
JoshNang | rloo: the endpoint direct_url returns can be an auth url in some cases, which won't work for directly downloading from with a temp url. it's something i want to make better in time, but i don't think requiring a swift endpoint url is unreasonable. here's the code explaining it better: https://github.com/openstack/glance/blob/5148c9648fc959c1d807313176afe3fcf84b89cf/glance/store/swift.py#L151 | 15:41 |
*** eghobo has joined #openstack-ironic | 15:41 | |
*** eghobo has quit IRC | 15:41 | |
*** eghobo has joined #openstack-ironic | 15:43 | |
*** athomas has joined #openstack-ironic | 15:43 | |
NobodyCam | brb | 15:43 |
*** eghobo has quit IRC | 15:43 | |
*** dkehn_ has joined #openstack-ironic | 15:45 | |
*** eghobo has joined #openstack-ironic | 15:45 | |
*** dkehnx has quit IRC | 15:49 | |
devananda | rloo_afk: after some mroe thought, +2'd | 15:50 |
*** Halacs has quit IRC | 15:50 | |
NobodyCam | lol got the recheck / reverify bug number all ready for 109033 | 15:53 |
NobodyCam | :( | 15:53 |
devananda | NobodyCam, Shrews: these are the reviews for BPs which we've targeted to J2 but have not landed yet -- https://etherpad.openstack.org/p/Du9hOAFg3d | 15:54 |
devananda | assuming the gate re-opens for us, it'd be great to get those landed today | 15:55 |
Shrews | devananda: ack | 15:55 |
NobodyCam | ack :) will do our best to get them in | 15:55 |
devananda | but, like, it's also just a milestone, so meh. don't stress about it. | 15:56 |
NobodyCam | lol /me still has to read the meeting notes from monday | 15:56 |
NobodyCam | :( | 15:56 |
devananda | and thanks! | 15:56 |
*** hemna has joined #openstack-ironic | 15:58 | |
dtantsur | devananda, https://review.openstack.org/#/c/86588/ actually could be approved already, just waiting for the gate to re-open | 16:02 |
NobodyCam | 109033 everifying now | 16:02 |
NobodyCam | reverifying even | 16:02 |
*** Mikhail_D_ltp has joined #openstack-ironic | 16:07 | |
devananda | dtantsur: wait a sec. https://review.openstack.org/#/c/86588/27/ironic/conductor/utils.py is removing the old support BEFORE porting SSH driver to use the new interface | 16:08 |
devananda | i think the patches are in the wrong order | 16:08 |
*** matty_dubs is now known as matty_dubs|lunch | 16:08 | |
dtantsur | devananda, did SSH ever had this support? | 16:08 |
*** dkehn_ is now known as dkehnx | 16:08 | |
dtantsur | devananda, judging by the SSH patch, no. So no need for old support there | 16:09 |
devananda | dtantsur: ah, I see. since SSH did not have a ManagementInterface before, it wouldnt error either, as that's not a requried interface | 16:11 |
*** foexle has joined #openstack-ironic | 16:17 | |
*** viktors is now known as viktors|afk | 16:20 | |
*** rloo_afk is now known as rloo | 16:22 | |
rloo | thx devananda ;) | 16:22 |
rloo | JoshNang: thx. It seems to me that if there's a direct url and the user hasn't specified the swift_endpoint_url, then maybe using the endpoint from the direct url might work. And it makes it more consistent with the other url-related configs, but I don't feel strongly about it. | 16:24 |
NobodyCam | brb | 16:24 |
JoshNang | rloo: i might work, or it might cause a frustrating to debug 401 error when trying to download the image. i think its safer to require it for now. i know someone was evaluating how to get the swift endpoint from keystone, which would make this all much easier | 16:28 |
rloo | JoshNang: fine with me then. | 16:29 |
Shrews | devananda: some of the Mgmt Iface patches need to have comments addressed before landing | 16:33 |
Shrews | rloo's patches are good to go today though | 16:33 |
*** jcoufal has quit IRC | 16:34 | |
devananda | mrda-away: one more patch for the nova driver -- https://review.openstack.org/#/c/109022/ | 16:34 |
devananda | mrda-away: I've annotated it on the etherpad, too | 16:34 |
devananda | Shrews: yea, it looks that way to me too. i just added some comments to https://etherpad.openstack.org/p/Du9hOAFg3d | 16:34 |
devananda | Shrews: those might not quite be ready yet | 16:34 |
rloo | Shrews: there was that one bug of Lucas' too that was for J2. sec. | 16:35 |
rloo | Shrews: https://review.openstack.org/100958 | 16:36 |
rloo | Shrews: I figured I'd wait for your comments first, before I looked again ;) | 16:36 |
Shrews | rloo: ah right. i'm satisfied by lucasagomes's comment. will re-review now. feel free to review it | 16:37 |
devananda | Shrews: I also don't see any client patch for lucas' management api | 16:37 |
lucasagomes | devananda, yeah I'm waiting the api to be merged | 16:37 |
lucasagomes | devananda, gotta do tempest patches as well | 16:37 |
lucasagomes | for the api | 16:37 |
lucasagomes | lemme take a look at it | 16:38 |
devananda | lucasagomes: right. so without a client change proposed yet, i'm going to bump that BP to J3 | 16:39 |
lucasagomes | devananda, ack | 16:39 |
lucasagomes | yeah makes sense | 16:39 |
lucasagomes | I will work on the client asap | 16:39 |
Shrews | lucasagomes: re 100958, shouldn't we be getting the _L* stuff from i18n.py now? | 16:40 |
lucasagomes | Shrews, oh, indeed we should | 16:40 |
devananda | gotta run to the airport, bbl | 16:40 |
lucasagomes | I just rebased | 16:40 |
lucasagomes | urgh | 16:40 |
lucasagomes | Shrews, will put a patch up quickly for that | 16:41 |
Shrews | lucasagomes: ok. +2 from me after that :) | 16:41 |
Shrews | devananda: have a good flight! | 16:41 |
lucasagomes | devananda, safe flight! | 16:41 |
lucasagomes | Shrews, :) ack | 16:41 |
*** jcoufal has joined #openstack-ironic | 16:43 | |
openstackgerrit | Lucas Alvares Gomes proposed a change to openstack/ironic: Fix nodes left in an incosistent state if no workers https://review.openstack.org/100958 | 16:45 |
lucasagomes | Shrews, ^ here we go :) | 16:45 |
Shrews | lucasagomes: awesome | 16:46 |
*** rainya has joined #openstack-ironic | 16:46 | |
*** rainya has quit IRC | 16:46 | |
*** rainya has joined #openstack-ironic | 16:47 | |
*** Alexei_9871 has quit IRC | 16:48 | |
*** derekh_ has quit IRC | 16:49 | |
*** jistr has joined #openstack-ironic | 16:50 | |
lucasagomes | Shrews, rloo you won't join us for pizza on sunday? | 16:51 |
rloo | lucasagomes: nope. Would like to, but am going to be home ;) | 16:51 |
Shrews | lucasagomes: i can't attend. :( went to the tripleo meeting this week instead | 16:52 |
rloo | lucasagomes: will have a croissant with you in Paris instead. | 16:52 |
lucasagomes | Shrews, rloo oh :/ | 16:52 |
lucasagomes | rloo, deal :D | 16:52 |
lucasagomes | gotta start practicing my french since now | 16:53 |
rloo | lucasagomes: some of us have to hold the fort down while you guys are partying ;) | 16:53 |
lucasagomes | lol | 16:53 |
*** rwsu has quit IRC | 16:58 | |
*** rwsu has joined #openstack-ironic | 16:59 | |
*** jdob has quit IRC | 17:01 | |
*** bvivek has joined #openstack-ironic | 17:02 | |
*** jdob has joined #openstack-ironic | 17:03 | |
*** matty_dubs|lunch is now known as matty_dubs | 17:05 | |
lucasagomes | aight, have a great night everyone | 17:05 |
lucasagomes | will call it a day, I still have to start packing my stuff and all | 17:05 |
*** Nisha has joined #openstack-ironic | 17:05 | |
*** lucasagomes is now known as lucas-dinner | 17:05 | |
NobodyCam | night lucas-dinner | 17:06 |
rloo | ciao lucas-dinner. | 17:07 |
rloo | adam_g: was wondering about 109022 and the reason for the patch. | 17:07 |
*** pelix has quit IRC | 17:07 | |
rloo | adam_g: I added a comment/my question to 109022. | 17:10 |
*** derekh_ has joined #openstack-ironic | 17:11 | |
*** marzif has quit IRC | 17:12 | |
*** wanyen has joined #openstack-ironic | 17:20 | |
*** lsmola has joined #openstack-ironic | 17:21 | |
Shrews | i wonder for the ipxe review if we should consider having an http_port option? | 17:23 |
JayF | Shrews: link? | 17:24 |
Shrews | JayF: https://review.openstack.org/99318 | 17:24 |
JayF | Shrews: I agree with you generally, but adding that config option was missed in the spec :( | 17:26 |
JayF | I'll comment on it with something saying I think we should anyway and see how it goes over :) | 17:26 |
Shrews | does that mean it can't be added now? | 17:26 |
NobodyCam | JayF: +! | 17:26 |
NobodyCam | +1 even | 17:26 |
NobodyCam | i would think its worth adding | 17:27 |
Shrews | i already left a comment | 17:27 |
Shrews | but feel free to add your own too | 17:27 |
JayF | Shrews: jinx :) | 17:29 |
jroll | devananda: poke me when you have a moment so we can talk about talks :) | 17:29 |
jroll | Shrews JayF, fwiw, I don't think anyone will kill that idea just because it isn't in the spec | 17:29 |
jroll | my thought is that should be a url eg. http://hostname:1234/somepath | 17:31 |
JayF | jroll: that doesn't match the usual way those vars are presented for other things though | 17:31 |
JayF | jroll: all other stuff is usually host/port/path separately | 17:31 |
jroll | yes it does | 17:31 |
JayF | jroll: although I personally agree that URLs are much better in most cases | 17:31 |
jroll | no, don't believe so | 17:31 |
jroll | pretty sure it is usually just an endpoint | 17:32 |
jroll | a la api_endpoint in configs | 17:32 |
JayF | heh, and it's a mix in the settings | 17:32 |
JayF | looks like keystoneclient can take both | 17:32 |
jroll | right | 17:33 |
JayF | and most other things use a url | 17:33 |
JayF | so I'd be very OK with that as well, but probably should've been mentioned in the spec :( | 17:33 |
jroll | but split options in keystone is deprecated :) | 17:33 |
jroll | agree, but | 17:33 |
* JayF doesn't want to crater lucas-dinner's already awesome work with something that should've been caught earlier | 17:33 | |
jroll | disagree with another spec for that | 17:33 |
jroll | do it or else i will :p | 17:33 |
JayF | I don't want 'another spec for that' | 17:33 |
jroll | and I hate reviewing on mobile | 17:34 |
JayF | I'm suggesting that something good shouldn't be held up for something better in a lot of cases | 17:34 |
rloo | FWIW, I don't think a spec is set in stone. | 17:34 |
jroll | if we land this as is someone will want a spec for the next patch | 17:34 |
JayF | jroll: and we should get good at telling $someone that they're wrong and being pedantic :P | 17:34 |
JayF | rloo: can we submit merge reqs to edit specs? | 17:34 |
JayF | rloo: I guess it's technically possible, pretty interesting when you think about design docs as code | 17:35 |
rloo | JayF: I was wondering about that. I'd like to. Guess that is something we can ask Deva. I'd like the specs to be kept up-to-date and clearly, a spec can't be totally 100% accurate even though we'd like it to be. | 17:35 |
*** harlowja has joined #openstack-ironic | 17:35 | |
rloo | JayF: otherwise, I'd have to ding JoshNang on his swift-url patch cuz it is a bit diff from the spec. | 17:35 |
jroll | right, but when does it end :/ | 17:36 |
jroll | w up to date specs | 17:36 |
rloo | jroll: so have you worked on a project where there were requirements/design docs, etc? | 17:36 |
rloo | jroll: where does it end there? seems like we're reasonable people (i'd like to think so). we should do what works for us so that we can do our best w/o a lot of overhead. | 17:37 |
jroll | rloo: kind of. I ran :) | 17:37 |
JayF | rloo: we all work in an environment where you're expected to make things work, write tests to prove it, then iterate over time. I've probably writen 4 design docs, tops, in my career, and they were mostly for things that couldn't be iterated upon (like setting up a DNS naming scheme) without a lot of pain | 17:37 |
rloo | caveat being 'except for what the openstack community imposes on us that we might disagree with ;) | 17:37 |
jroll | rloo: what I mean is... if ipxe code changes over time, do we keep updating the spec etx | 17:38 |
jroll | idk if I agree with updating the spec | 17:38 |
rloo | jroll: ahhh. that's a good question. I think that depends on how-much the ipxe code changes. | 17:38 |
jroll | especially because we're now freezing new ironic-specs patches :p | 17:39 |
JayF | I guess here is the core question: | 17:39 |
JayF | - Should people who aren't Openstack Developers use specs for anything at all ever | 17:39 |
JayF | i.e. they aren't supposed to be documentation for anyone but the people implementing it, right? | 17:39 |
openstackgerrit | Nisha Agarwal proposed a change to openstack/ironic-specs: iLO Node discovery https://review.openstack.org/109088 | 17:39 |
rloo | JayF: seems like I saw something in the list about that, and no, the specs are for developers if i recall correctly. | 17:39 |
NobodyCam | lol love it when sales men call with info that is YEARS out of date | 17:39 |
JayF | rloo: in that case I agree with Jim, that we shouldn't /need/ to update specs in 99.999% of cases | 17:40 |
jroll | JayF: thing is, PMs etc use specs | 17:40 |
jroll | also release notes | 17:40 |
jroll | etc | 17:41 |
jroll | annoying but true | 17:41 |
JayF | the .001% of specs that would warrant a change would be the ones that the substance of the feature changed | 17:41 |
JayF | i.e. if someone is using a spec to configure 'http_host" for iPXE, they're probably going to be in for a bad time | 17:41 |
rloo | what's a PM? | 17:41 |
*** jistr has quit IRC | 17:42 | |
JayF | but overall, even if the option structure changes a little, it's still essentially the same | 17:42 |
JayF | rloo: Product Manager || Project Manager | 17:42 |
rloo | we don't have any in openstack, do we? | 17:42 |
JayF | Companies that sponsor Openstack development have lots and lots of 'em :) | 17:42 |
rloo | JayF: well don't tell your PMs about the specs then ;) | 17:43 |
JayF | rloo: I don't think any of our PMs use them at all | 17:43 |
JayF | rloo: if anything; we err too much on the other side -- internal docs and conversations about features that we know will end up upstream | 17:43 |
* Shrews notes that asking even the simplest of questions can lead to much IRC scrollback | 17:44 | |
rloo | Shrews - I think to summarize, the question probably for devananda, is when/if a spec should be updated after it has merged. | 17:46 |
rloo | oh, and is it OK to deviate from the spec, eg different/more configs. | 17:46 |
adam_g | rloo, re: 109022 we need to ensure nodes are cleaned up after all exceptions, not only the ones we're currently catching | 17:47 |
rloo | adam_g: just for spawning or all cases? | 17:47 |
*** rainya has quit IRC | 17:49 | |
adam_g | rloo, spawning, so that failed provisions properly disassociate instances from nodes and kick resources back to nova | 17:49 |
rloo | adam_g: ok. and is it ok to reraise ironic exceptions? | 17:50 |
adam_g | rloo, failures elsewhere usually cause things to stay in some ERROR state waiting for admin action (ie, rebuild) | 17:50 |
devananda | rloo: that's an evolving discussion | 17:51 |
devananda | ideally, i think we shouldn't approve a spec until we're comfortable with, and have thought through, the implications, required options, etc | 17:51 |
devananda | updating a spec after landing it means the design was wrong / incomplete. but also, we admit mistakes and learn from them, so maybe that's ok? | 17:52 |
devananda | i don't have a cut-anbd-dry view on that yet | 17:52 |
lucas-dinner | Shrews, JayF yeah port is useful... the way I was configuring the port was to put it as part of the http_server value | 17:53 |
rloo | devananda: yes, I think it's OK and I'd prefer that it be updated than for it to be mis-leading or whatever. | 17:53 |
lucas-dinner | like https://review.openstack.org/#/c/99677/9/lib/ironic (L265) | 17:53 |
lucas-dinner | Shrews, JayF but it's not very intuitive... I will add a config option for port there | 17:53 |
JayF | lucas-dinner: great! Thanks! | 17:53 |
JayF | lucas-dinner: jroll was suggesting it change to just a fully qualified URL, which I'm personally OK with as well, but either is fine :) | 17:54 |
rloo | devananda: otherwise, I worry that we'd spend even more time over the specs than we do now and I'm not sure we (well I) want to do that. | 17:54 |
* JayF just glad iPXE support is going in mainline | 17:54 | |
adam_g | rloo, re-raising the non-nova exceptions should be fine, the compute manager will catch those and cause a reschedule | 17:54 |
lucas-dinner | JayF, :D right... jroll if u want please leave a comment on the patch and I will address it tomorrow asap | 17:54 |
rloo | adam_g: as long as it is ok with nova to expect non-nova exceptions. eg in ironic, we don't want non-ironic exceptions to make it 'up' the chain. | 17:55 |
adam_g | rloo, https://git.openstack.org/cgit/openstack/nova/tree/nova/compute/manager.py#n1344 | 17:55 |
* lucas-dinner brb again | 17:55 | |
rloo | adam_g: so nova uses the big Exception hammer to make sure they get it all. | 17:56 |
rloo | adam_g: thx. i'll take another look at that patch. | 17:57 |
adam_g | rloo, np :) | 17:57 |
adam_g | Shrews, | 17:57 |
adam_g | Run: 663 in 488.563237 sec. | 17:57 |
adam_g | - Passed: 600 | 17:57 |
adam_g | - Skipped: 63 | 17:57 |
adam_g | - Failed: 0 | 17:57 |
adam_g | thats smoke-serial minus scenarios | 17:58 |
stevebaker | devananda: hey, https://review.openstack.org/#/c/92257/ is ready for another look | 18:07 |
*** Nisha has quit IRC | 18:07 | |
devananda | adam_g: awesome. that all updated on the 'pad, I presume? | 18:07 |
adam_g | devananda, yup. required patches @ https://etherpad.openstack.org/p/IronicCI | 18:08 |
devananda | sweet | 18:08 |
devananda | i'll poke folks next week in portland, if they aren't already landed | 18:08 |
openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding swift temp url support https://review.openstack.org/81391 | 18:09 |
devananda | stevebaker: LGTM, but please don't merge it until it passes check-tempest-dsvm-virtual-ironic-nv | 18:10 |
devananda | that's non-voting so you need to atually look at it, not just the jenkins vote | 18:11 |
* devananda boards a plane | 18:11 | |
Shrews | adam_g: that's waaay more tests than what my test run had | 18:11 |
stevebaker | devananda: ok. are there known other failures on that job? | 18:11 |
* Shrews scratches head | 18:11 | |
devananda | stevebaker: only the gate problem which nova created yesterday | 18:11 |
devananda | stevebaker: once that merges the job SHOULd work again | 18:11 |
* devananda really, actually boards a plane | 18:12 | |
stevebaker | devananda: ok. | 18:12 |
stevebaker | devananda: See you in Hawaii ;) | 18:12 |
Shrews | adam_g: my notes are at the bottom of https://etherpad.openstack.org/p/IronicTempestFeatures | 18:13 |
NobodyCam | brb | 18:15 |
* rloo wonders why we can't have an OpenStack summit or midcycle meetup in Hawaii. | 18:17 | |
NobodyCam | rloo: ++ | 18:21 |
* GheRivero will never take a >26 hours flight just to go to Hawaii just for work | 18:21 | |
NobodyCam | GheRivero: ask your "better half" | 18:22 |
NobodyCam | lol my bet is they would want to go | 18:22 |
rloo | GheRivero: yeah, don't do it "just for work"! Is it really that long a flight? | 18:23 |
GheRivero | rloo: +2h for check in and another 2.30h ride to the airport... so it's actually worst | 18:26 |
NobodyCam | GheRivero: :( | 18:26 |
rloo | GheRivero: but you love where you live, right? :D Anyway, I seriously doubt we'll be going to Hawaii. | 18:27 |
*** amitpp has joined #openstack-ironic | 18:27 | |
GheRivero | :) I can always stay an extra month for holidays :) | 18:28 |
* NobodyCam would be happy with a extra week | 18:29 | |
NobodyCam | :) | 18:29 |
rloo | GheRivero: sweet! | 18:32 |
*** rainya has joined #openstack-ironic | 18:37 | |
*** bvivek has quit IRC | 18:38 | |
NobodyCam | 109033 almost there. two tests | 18:39 |
NobodyCam | left | 18:39 |
GheRivero | \o/ | 18:41 |
*** amitpp has quit IRC | 18:43 | |
NobodyCam | update all test on 109033 : Success! | 18:55 |
jroll | I want to propose a spec to add a "notes" field to the node table... but that's so cmdb | 18:58 |
jroll | or like make maintenance a text field :) | 18:58 |
NobodyCam | oh that is very cmdb like | 19:00 |
jroll | I know | 19:01 |
jroll | but like... | 19:01 |
jroll | so useful | 19:02 |
NobodyCam | brb | 19:03 |
rloo | how long do we wait after the merge, to do rechecks? | 19:07 |
* rloo tries a recheck bug 1347795 and crosses fingers | 19:09 | |
*** lsmola has quit IRC | 19:20 | |
greghaynes | Has there been any thought on giving conductors a uuid in the schema? Curious if this has just not been needed or is intentionally not done due to conductor.hostname? | 19:26 |
greghaynes | I need to add a filter to nodes based on conductor for some context.... | 19:26 |
openstackgerrit | Adam Gandelman proposed a change to openstack/ironic: Generalize exception handling in Nova driver https://review.openstack.org/109022 | 19:27 |
lifeless | greghaynes: you do? | 19:28 |
lifeless | greghaynes: the hostname is the key to use | 19:28 |
greghaynes | lifeless: ok. devananda mentioned that for the hashring rebalance it might be better to rebalance nodes filtered by the conductor that is now missing which could be a value stored in the db | 19:29 |
greghaynes | rather than walking the whole list of nodes | 19:29 |
lifeless | yes, thats a query on the hostname | 19:30 |
lifeless | you want the current owning conductor to be a field on the node | 19:30 |
lifeless | then its select from nodes where owner=failed_conductor | 19:30 |
greghaynes | Yep, that should work fine | 19:30 |
*** scubacuda has joined #openstack-ironic | 19:55 | |
*** romcheg1 has quit IRC | 20:01 | |
*** ChuckC has quit IRC | 20:07 | |
*** chuckC has joined #openstack-ironic | 20:13 | |
*** chuckC has quit IRC | 20:14 | |
*** marzif has joined #openstack-ironic | 20:19 | |
rloo | anyone out there, NobodyCam, Shrews, this one might have a chance of making it in today if you have time: https://review.openstack.org/#/c/109022/. | 20:26 |
rloo | although I guess deva should still be the one that approves. | 20:26 |
Shrews | rloo: yeah, that's deva's call for driver related patches | 20:26 |
rloo | still need eyeballs on it though. | 20:27 |
Shrews | rloo: i'll look it over | 20:31 |
rloo | adam_g thanks you Shrews :-) | 20:31 |
Shrews | rloo: thx for handling the rechecks | 20:35 |
rloo | yw Shrews. Still crossing my fingers... | 20:35 |
* Shrews trying to simultaneously review code and deal with getting his door replaced after an attempted break in | 20:36 | |
rloo | Shrews: not good, the door is more important than the review. | 20:36 |
devananda | i'm in the airport for another hour or two, apparently | 20:41 |
devananda | jroll: what are yall doing for network isolation of tenants? | 20:42 |
openstackgerrit | A change was merged to openstack/python-ironicclient: Add driver-properties command https://review.openstack.org/76338 | 20:43 |
devananda | woot! our gate's workign! | 20:43 |
rloo | whoohoo! I think I waited 15-30 min or so (it's all a blur) before rechecking the others. So there should be a few more merges soon/today, unless there are conflicts. | 20:44 |
devananda | awesome, thanks | 20:44 |
rloo | Shrews, devananda beat you to 109022. (Go take care of the door.) | 20:49 |
Shrews | dang it devananda. quit approving patches i'm reviewing! | 20:49 |
Shrews | :-P | 20:49 |
devananda | lol | 20:49 |
* Shrews +2's out of spite | 20:50 | |
Shrews | :) | 20:50 |
devananda | Shrews: all yours. i'll go back to writing my slides for tomorrow | 20:50 |
Shrews | devananda: in rtp still? | 20:51 |
Shrews | err, rdu | 20:51 |
devananda | ya | 20:51 |
devananda | rerouted. ATL is hall thunderstorms | 20:51 |
NobodyCam | oh happy happy joy joy | 20:52 |
devananda | JayF: ping since jroll didn't reply yet | 20:52 |
Shrews | devananda: ah. so, one very important thing i learned about traveling living in this area... NEVER route through ATL in the summer during evening hours | 20:52 |
devananda | Shrews: good to know. | 20:52 |
Shrews | you're almost guaranteed to be delayed due to weather | 20:52 |
Shrews | i try to route through MSP whenever possible (for Delta, at least) | 20:54 |
JayF | devananda: I saw the question, and I 1) Don't understand 100% all the technical things about how we do it and 2) Don't know how much of it we are making public | 20:55 |
JayF | devananda: it boils down to $vendor_things, and I assure you we tested to ensure the network isolation (i.e., no arp poisoning or IP-stealing) worked properly | 20:55 |
devananda | JayF: cool. that confirms my thinking that you *are* doing it. | 20:56 |
JayF | devananda: yes, we absolutely are :) | 20:56 |
*** mrda-away is now known as mrda | 21:01 | |
mrda | Morning Ironic! | 21:01 |
NobodyCam | morening mrda | 21:03 |
mrda | NobodyCam: o/ | 21:04 |
mrda | devananda: yay, 107882 approved! | 21:05 |
*** cdent has quit IRC | 21:07 | |
*** derekh_ has quit IRC | 21:08 | |
* devananda wanders off to get a new boarding pass | 21:13 | |
NobodyCam | brb | 21:28 |
openstackgerrit | A change was merged to openstack/ironic: Generalize exception handling in Nova driver https://review.openstack.org/109022 | 21:29 |
rloo | mrda: do you have time to rebase 107882? | 21:32 |
mrda | of course! | 21:32 |
mrda | :) | 21:33 |
* NobodyCam thinks he's comming down with some form of travel sickness :( | 21:34 | |
openstackgerrit | A change was merged to openstack/ironic: Backporting nova host manager changes into ironic https://review.openstack.org/108872 | 21:35 |
rloo | NobodyCam: go home and stay home! | 21:35 |
* NobodyCam is home... lol and am here until sunday lol | 21:36 | |
*** Mikhail_D_ltp has quit IRC | 21:36 | |
rloo | Oh, didn't realize you were home already NobodyCam. Well, you have a few days to recuperate :D | 21:36 |
NobodyCam | yep.. :) | 21:37 |
openstackgerrit | A change was merged to openstack/ironic: Implement API to get driver properties https://review.openstack.org/107096 | 21:40 |
*** jdob has quit IRC | 21:43 | |
*** rwsu has quit IRC | 21:44 | |
*** jcoufal has quit IRC | 21:44 | |
openstackgerrit | Michael Davies proposed a change to openstack/ironic: Import fixes from the Nova driver reviews https://review.openstack.org/107882 | 21:51 |
*** matty_dubs is now known as matty_dubs|gone | 21:52 | |
rloo | I see a failure in check-tempest-dsvm-virtual-ironic, dib-run-parts, "rmdir: failed to remove '/tmp/image.JgwLksHL/mnt/tmp/in_target.d': Device or resource busy" | 21:58 |
rloo | http://logs.openstack.org/88/86588/27/check/check-tempest-dsvm-virtual-ironic/ac170fe/logs/devstacklog.txt.gz | 21:58 |
openstackgerrit | A change was merged to openstack/ironic: Catch oslo.db error instead of sqlalchemy error https://review.openstack.org/108943 | 21:59 |
rloo | guess what I'm going to do... | 21:59 |
NobodyCam | lol file a bug? | 22:00 |
NobodyCam | :-p | 22:00 |
*** dguerri is now known as _dguerri | 22:00 | |
rloo | something even easier NobodyCam, but I won't mention it here. | 22:00 |
NobodyCam | lol | 22:01 |
NobodyCam | shhh | 22:01 |
* NobodyCam wont tell | 22:01 | |
rloo | heh. I think everything else has merged, except for that one failure and the rebase. | 22:01 |
openstackgerrit | A change was merged to openstack/ironic: Fix nodes left in an incosistent state if no workers https://review.openstack.org/100958 | 22:02 |
rloo | 107882 needed to be rebased. Could someone +2/approve? devananda already approved it once. | 22:04 |
rloo | oh, maybe I can do that myself. | 22:04 |
* mrda is off to propose new patches to nova, now that things have merged on our side | 22:06 | |
openstackgerrit | Josh Gachnang proposed a change to openstack/ironic: Adding swift temp url support https://review.openstack.org/81391 | 22:08 |
*** chuckC has joined #openstack-ironic | 22:12 | |
rloo | JoshNang: thx for updating 81391. I have to take off now, most likely won't get around to looking again until next Mon. I hope it'll be approved before then ;) | 22:13 |
*** marzif has quit IRC | 22:16 | |
JoshNang | rloo: no worries! thanks for the great reviews! | 22:29 |
*** jgrimm has quit IRC | 22:39 | |
*** chuckC has quit IRC | 22:47 | |
*** chuckC has joined #openstack-ironic | 22:53 | |
*** radsy has joined #openstack-ironic | 22:54 | |
*** rloo has quit IRC | 22:56 | |
*** chuckC has quit IRC | 23:09 | |
openstackgerrit | A change was merged to openstack/ironic: Import fixes from the Nova driver reviews https://review.openstack.org/107882 | 23:14 |
openstackgerrit | A change was merged to openstack/ironic: IPMINative to use the new ManagementInterface https://review.openstack.org/86588 | 23:19 |
*** athomas has quit IRC | 23:20 | |
*** lucas-dinner has quit IRC | 23:28 | |
*** wanyen has quit IRC | 23:31 | |
*** jcoufal has joined #openstack-ironic | 23:45 | |
mrda | So, pxe_bootfile_name config option in the nova ironic driver... | 23:50 |
mrda | it's currently in our tree, and is referenced directly, or indirectly in the following files: | 23:50 |
mrda | ironic/common/neutron.py ironic/common/pxe_utils.py ironic/drivers/modules/pxe.py ironic/tests/test_neutron.py ironic/tests/test_pxe_utils.py | 23:51 |
mrda | If I remove these references, all tests pass - meaning everything is ok, right? :) | 23:51 |
mrda | Any objections before I put up this patch for review? | 23:51 |
jroll | ummmmmmmmmm | 23:52 |
jroll | did you try tempest? | 23:52 |
jroll | that has to fail horribly | 23:52 |
mrda | ok, thanks jroll | 23:52 |
jroll | mrda: just my guess, I know nothing :) | 23:52 |
jroll | put it up and see what happens | 23:52 |
mrda | I have this comment from Deva on this in https://review.openstack.org/#/c/103167/7/nova/virt/ironic/driver.py "This has been refactored out, and isn't needed here any more" | 23:52 |
jroll | and watch deva strike it down | 23:53 |
mrda | So I'm not a pxe expert, but it appears it's not needed in the nova driver for ironic - just wondering if it's needed internally | 23:53 |
jroll | hm, you might be right | 23:54 |
jroll | I'm not a pxe expert either :P | 23:54 |
mrda | ok, here comes the patch - wheee! | 23:54 |
* jroll braces himself | 23:55 | |
mrda | lol | 23:58 |
openstackgerrit | Michael Davies proposed a change to openstack/ironic: Remove config option pxe_bootfile_name & vestiges https://review.openstack.org/109455 | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!