*** vint_bra has quit IRC | 00:05 | |
clayg | briancline: nope! | 00:30 |
---|---|---|
*** itlinux has joined #openstack-swift | 00:45 | |
*** itlinux has quit IRC | 00:51 | |
briancline | clayg: I've probably been working for too long today but pretty sure I found one in production :| | 00:52 |
*** bigdogstl has joined #openstack-swift | 00:52 | |
briancline | or, possibly more likely, some convoluted series of events led to data being put into a place where it shouldn't be | 00:52 |
clayg | everytime I think about "improbable" I *feel* like that means *possible* - but then I do some math - and I can't seem to convince myself the entropy of the universe has been exhausted. It's weird state of suspended disbelief. | 00:55 |
*** tovin07_ has joined #openstack-swift | 00:56 | |
clayg | briancline: there's this -> https://github.com/openstack/swift/blob/master/swift/obj/diskfile.py#L2307 | 00:56 |
*** bigdogstl has quit IRC | 00:56 | |
clayg | which at least says if there is an existing object '/a/c/o' at has xyz, then later another object '/a2/c/o' is directed into that same hashdir the diskfile just barfs on itself. | 00:57 |
clayg | I always assumed it was not possible unless you have access to the swift.conf hash secrets - and bunch of GPU and superadmin write access to any possible collision path (a preimage attack would be significantly more work - and limited to data w/i your same own account) | 00:59 |
clayg | *possible to hit that code path | 00:59 |
briancline | clayg: yeah, the object server catches that exception and throws a 403, which is exactly what I'm seeing | 01:00 |
clayg | briancline: oic! | 01:01 |
briancline | which is good, it at least hits the brakes on the request, just made me go "huh?" seeing a 403 in object server logs | 01:01 |
clayg | if you swift-object-info that object *in* that hashdir does the location information indicate it should be somewhere else entirely? | 01:01 |
briancline | for a colliding PUT, the proxy ends up writing it to handoffs, which accept it | 01:02 |
clayg | oh neat! then rsync hates you :'( | 01:02 |
briancline | hm lemme check. i did notice swift-get-nodes for the same path returns some strange disparities | 01:02 |
clayg | well... gotta bounce | 01:03 |
*** chsc has quit IRC | 01:04 | |
*** _ix has joined #openstack-swift | 01:11 | |
*** bigdogstl has joined #openstack-swift | 01:11 | |
*** _ix has quit IRC | 01:17 | |
*** portante has quit IRC | 01:17 | |
*** portante has joined #openstack-swift | 01:18 | |
*** bigdogstl has quit IRC | 01:20 | |
*** kei_yama has quit IRC | 01:31 | |
openstackgerrit | Siyi Luo proposed openstack/python-swiftclient master: Drop py34 target in tox.ini and setup.cfg https://review.openstack.org/506431 | 01:43 |
*** ukaynar has quit IRC | 01:55 | |
*** ukaynar has joined #openstack-swift | 01:55 | |
*** ukaynar_ has joined #openstack-swift | 01:59 | |
*** ukaynar has quit IRC | 01:59 | |
*** gkadam has quit IRC | 01:59 | |
*** mingyu has joined #openstack-swift | 01:59 | |
*** kei_yama has joined #openstack-swift | 02:08 | |
*** vint_bra has joined #openstack-swift | 02:10 | |
*** bigdogstl has joined #openstack-swift | 02:16 | |
*** bigdogstl has quit IRC | 02:21 | |
notmyname | tdasilva: FWIW, I've got a server and public IP set up to host the LP bug stats. now I need to install the software and choose a domain name to point at it | 02:24 |
tdasilva | notmyname: cool | 02:24 |
tdasilva | are you still using your dashboard? | 02:25 |
notmyname | tdasilva: sortof | 02:28 |
*** ukaynar_ has quit IRC | 02:46 | |
*** vint_bra has quit IRC | 02:52 | |
openstackgerrit | Merged openstack/swift feature/deep: Address some lingering review comments https://review.openstack.org/506367 | 02:57 |
*** zaitcev has quit IRC | 03:08 | |
*** zaitcev has joined #openstack-swift | 03:11 | |
*** ChanServ sets mode: +v zaitcev | 03:11 | |
*** gkadam has joined #openstack-swift | 03:13 | |
*** MalcomX36 has joined #openstack-swift | 03:20 | |
*** bigdogstl has joined #openstack-swift | 03:35 | |
*** bigdogstl has quit IRC | 03:42 | |
*** psachin has joined #openstack-swift | 03:47 | |
mattoliverau | notmyname: bugs.not.mn? | 03:55 |
notmyname | maybe | 03:56 |
*** links has joined #openstack-swift | 03:57 | |
mattoliverau | bugs.swiftstats.org then you can have dash.swiftstats.org for the dashboard :) | 03:59 |
*** Sukhdev has joined #openstack-swift | 04:22 | |
*** klrmn has quit IRC | 04:30 | |
*** itlinux has joined #openstack-swift | 04:46 | |
*** psachin has quit IRC | 04:48 | |
*** itlinux has quit IRC | 04:49 | |
*** itlinux has joined #openstack-swift | 04:51 | |
*** gyee has quit IRC | 04:55 | |
*** psachin has joined #openstack-swift | 05:03 | |
*** tone_z has quit IRC | 05:11 | |
*** itlinux has quit IRC | 05:12 | |
*** Sukhdev has quit IRC | 05:27 | |
*** shaner has quit IRC | 05:29 | |
*** shaner has joined #openstack-swift | 05:30 | |
*** psachin has quit IRC | 05:32 | |
*** psachin has joined #openstack-swift | 05:37 | |
*** bigdogstl has joined #openstack-swift | 05:39 | |
*** bigdogstl has quit IRC | 05:43 | |
*** skudlik has joined #openstack-swift | 05:46 | |
*** skudlik has left #openstack-swift | 05:47 | |
*** ChubYann has quit IRC | 05:49 | |
*** geaaru has joined #openstack-swift | 05:51 | |
*** shaner has quit IRC | 06:06 | |
*** shaner has joined #openstack-swift | 06:11 | |
*** SkyRocknRoll has joined #openstack-swift | 06:18 | |
*** mahatic has quit IRC | 06:25 | |
*** rcernin has joined #openstack-swift | 06:56 | |
*** cbartz has joined #openstack-swift | 06:58 | |
*** bigdogstl has joined #openstack-swift | 07:00 | |
*** bigdogstl has quit IRC | 07:02 | |
*** tesseract has joined #openstack-swift | 07:11 | |
*** bigdogstl has joined #openstack-swift | 07:12 | |
openstackgerrit | Tovin Seven proposed openstack/swift master: OSprofiler in OpenStack Swift https://review.openstack.org/468316 | 07:15 |
*** bigdogstl has quit IRC | 07:16 | |
*** kei_yama has quit IRC | 07:26 | |
*** pcaruana has joined #openstack-swift | 07:34 | |
*** kei_yama has joined #openstack-swift | 07:47 | |
*** kota_ has quit IRC | 08:04 | |
*** spotz has quit IRC | 08:05 | |
*** tanee has quit IRC | 08:07 | |
*** timur has quit IRC | 08:07 | |
*** timur has joined #openstack-swift | 08:08 | |
*** kota_ has joined #openstack-swift | 08:09 | |
*** ChanServ sets mode: +v kota_ | 08:09 | |
*** spotz has joined #openstack-swift | 08:10 | |
*** tanee has joined #openstack-swift | 08:14 | |
*** bigdogstl has joined #openstack-swift | 08:53 | |
*** bigdogstl has quit IRC | 08:58 | |
*** mingyu has quit IRC | 09:04 | |
*** kei_yama has quit IRC | 09:17 | |
*** early has quit IRC | 09:53 | |
acoles | timburke: +1 for get_own_shard_range() - I'd been trying to come up with a distinctive name for that method | 09:54 |
*** early has joined #openstack-swift | 09:55 | |
*** SkyRocknRoll has quit IRC | 09:56 | |
*** tovin07_ has quit IRC | 10:01 | |
*** psachin has quit IRC | 10:04 | |
*** mingyu has joined #openstack-swift | 10:04 | |
*** bigdogstl has joined #openstack-swift | 10:05 | |
*** mingyu has quit IRC | 10:09 | |
*** bigdogstl has quit IRC | 10:09 | |
*** chlong has quit IRC | 10:16 | |
openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: sharding: make find_shard_ranges a broker method https://review.openstack.org/505763 | 10:22 |
*** psachin has joined #openstack-swift | 10:23 | |
*** vint_bra has joined #openstack-swift | 10:27 | |
*** d0ugal has quit IRC | 10:48 | |
*** d0ugal has joined #openstack-swift | 10:48 | |
*** d0ugal has joined #openstack-swift | 10:48 | |
*** d0ugal has quit IRC | 10:48 | |
*** links has quit IRC | 11:03 | |
*** bigdogstl has joined #openstack-swift | 11:06 | |
*** bigdogstl has quit IRC | 11:10 | |
*** mingyu has joined #openstack-swift | 11:15 | |
*** links has joined #openstack-swift | 11:21 | |
*** links has quit IRC | 11:29 | |
*** links has joined #openstack-swift | 11:42 | |
*** MVenesio has joined #openstack-swift | 11:51 | |
*** vint_bra has quit IRC | 11:59 | |
*** bigdogstl has joined #openstack-swift | 12:06 | |
*** bigdogstl has quit IRC | 12:11 | |
*** silor has joined #openstack-swift | 12:13 | |
*** gkadam has quit IRC | 12:31 | |
*** catintheroof has joined #openstack-swift | 12:50 | |
*** bigdogstl has joined #openstack-swift | 13:07 | |
*** bigdogstl has quit IRC | 13:12 | |
*** links has quit IRC | 13:28 | |
*** gordc has joined #openstack-swift | 13:29 | |
gordc | hi folks, was wondering if something changed in swift master recently that requires any deployment changes? we use pifpaf to test against swift and seeing this now: https://github.com/jd/pifpaf/issues/77 | 13:31 |
gordc | in case youre interested, this is how we start swift: https://github.com/jd/pifpaf/blob/master/pifpaf/drivers/swift.py#L117-L132 | 13:32 |
*** vint_bra has joined #openstack-swift | 13:36 | |
*** Dinesh_Bhor has quit IRC | 13:36 | |
jrichli | gordc: is the 'fakedisk' with insufficient storage supposed to be used here? were you running a failure test case? | 13:38 |
tdasilva | gordc: I think you might be running into the same issue cschwede fixed recently for tripleo: https://bugs.launchpad.net/tripleo/+bug/1718454 | 13:40 |
openstack | Launchpad bug 1718454 in tripleo "undercloud, swift returns 503 in undercloud install" [Critical,Fix released] - Assigned to Christian Schwede (cschwede) | 13:40 |
openstackgerrit | Merged openstack/swift feature/deep: Add test for reversed listings https://review.openstack.org/505784 | 13:40 |
gordc | jrichli: nope. pifpaf is basically just temporarily setups and runs a service so it can be tested against | 13:40 |
acoles | gordc: looks like mount check is failing for account HEAD, this did change recently https://review.openstack.org/#/c/466255/ | 13:41 |
patchbot | patch 466255 - swift - Make mount_check option usable in containerized en... (MERGED) | 13:41 |
acoles | but looks like the change should help not fail | 13:41 |
cschwede | gordc: tdasilva is right, looks like the same issue we had in tripleo | 13:41 |
gordc | tdasilva: that does look quite similar. i'll take a look at it | 13:41 |
gordc | thanks for the pointers everyone, a lot more progress than me just looking blindly at code :) | 13:41 |
cschwede | gordc: see also this lp entry: https://launchpad.net/bugs/1718454 | 13:42 |
openstack | Launchpad bug 1718454 in tripleo "undercloud, swift returns 503 in undercloud install" [Critical,Fix released] - Assigned to Christian Schwede (cschwede) | 13:42 |
cschwede | and this is the change in swift that changed: https://github.com/openstack/swift/commit/30fd4e | 13:42 |
cschwede | acoles: ^^ it's the follow up that merged a few days ago, which broke quite a few CIs... | 13:43 |
gordc | cschwede: if i create folder, does it work for older versions of swift too? | 13:43 |
cschwede | gordc: yes | 13:43 |
acoles | cschwede: ah, glad your are on the case :) | 13:43 |
acoles | you are* | 13:43 |
gordc | cschwede: cool cool | 13:44 |
*** MVenesio has quit IRC | 14:08 | |
*** psachin has quit IRC | 14:32 | |
*** chlong has joined #openstack-swift | 14:37 | |
*** catintheroof has quit IRC | 14:42 | |
*** hseipp has joined #openstack-swift | 14:43 | |
*** hseipp has quit IRC | 14:43 | |
*** hseipp has joined #openstack-swift | 14:43 | |
openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Make broker.build_shard_ranges filter results for marker etc https://review.openstack.org/506681 | 14:50 |
*** gordc has left #openstack-swift | 15:01 | |
*** chsc has joined #openstack-swift | 15:01 | |
*** chsc has joined #openstack-swift | 15:01 | |
*** Chinkara has joined #openstack-swift | 15:02 | |
*** klrmn has joined #openstack-swift | 15:05 | |
*** Chinkara has quit IRC | 15:07 | |
*** bigdogstl has joined #openstack-swift | 15:09 | |
*** Chinkara has joined #openstack-swift | 15:10 | |
*** bigdogstl has quit IRC | 15:14 | |
*** gyee has joined #openstack-swift | 15:36 | |
*** links has joined #openstack-swift | 15:36 | |
notmyname | good morning | 15:41 |
timburke | good morning | 15:47 |
*** Chinkara has quit IRC | 15:48 | |
timburke | hey, jrichli's here! i feel like i haven't heard from you in a while :-) | 15:51 |
jrichli | timburke: yeah, I haven't done much swift in a long time. Things keep coming up. | 15:52 |
notmyname | hello jrichli! | 15:53 |
notmyname | missed you in Denver | 15:53 |
jrichli | i missed being there! | 15:53 |
*** cbartz has quit IRC | 15:55 | |
*** hseipp has quit IRC | 16:06 | |
*** klrmn has quit IRC | 16:08 | |
*** hseipp has joined #openstack-swift | 16:10 | |
*** Chinkara has joined #openstack-swift | 16:11 | |
*** hseipp has quit IRC | 16:13 | |
*** hseipp has joined #openstack-swift | 16:13 | |
*** Chinkara has quit IRC | 16:14 | |
*** links has quit IRC | 16:21 | |
*** psachin has joined #openstack-swift | 16:23 | |
acoles | jrichli: curiously we drank much less beer in your absence ;-) | 16:23 |
openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Make broker.build_shard_ranges filter results for marker etc https://review.openstack.org/506681 | 16:23 |
*** hseipp has quit IRC | 16:27 | |
*** juzuluag has joined #openstack-swift | 16:28 | |
*** bigdogstl has joined #openstack-swift | 16:31 | |
jrichli | acoles: lol - are you not stepping up as party master? | 16:34 |
acoles | jrichli: nah - leave it to the youngsters | 16:35 |
*** bigdogstl has quit IRC | 16:35 | |
jrichli | tdasilva can handle it | 16:35 |
jrichli | you all have a good weekend. i gotta catch a flight | 16:36 |
acoles | jrichli: you too | 16:37 |
timburke | bye jrichli! thanks for checking in | 16:37 |
openstackgerrit | Alistair Coles proposed openstack/swift master: Follow-up test fixup https://review.openstack.org/505454 | 16:46 |
*** klrmn has joined #openstack-swift | 17:02 | |
*** bigdogstl has joined #openstack-swift | 17:04 | |
*** bigdogstl has quit IRC | 17:15 | |
*** bigdogstl has joined #openstack-swift | 17:16 | |
*** juzuluag has quit IRC | 17:17 | |
tdasilva | zaitcev, torgomatic, notmyname, clayg, acoles: I'm scheduling the meeting about p 427911 for Tuesday at 18:00 UTC | 17:18 |
patchbot | https://review.openstack.org/#/c/427911/ - swift - PUT+POST and its development test | 17:18 |
tdasilva | cschwede: ^ | 17:18 |
acoles | tdasilva: ok thanks | 17:18 |
tdasilva | seems to be the best time according to https://beta.doodle.com/poll/mfcd2ypt28fmce8y | 17:18 |
notmyname | I think that's 11am pacific | 17:19 |
clayg | notmyname: that's what I came up with (cc @tburke if you want) | 17:19 |
tdasilva | notmyname, clayg sounds right | 17:20 |
clayg | tdasilva: do you have A/V sorted? | 17:20 |
tdasilva | I can forward a meeting invite | 17:20 |
tdasilva | clayg: yeah | 17:20 |
clayg | i just need a link or something... | 17:21 |
timburke | yeah, i'd be somewhat interested. sorry for not responding to the poll, but i knew PDT would be well-represented :-) | 17:21 |
tdasilva | clayg: https://bluejeans.com/704762476 | 17:21 |
clayg | my meeting invite already has a room on it | 17:21 |
*** itlinux has joined #openstack-swift | 17:22 | |
*** hoonetorg has quit IRC | 17:22 | |
clayg | tdasilva: you're kicking ass - nice work | 17:25 |
*** mvk has quit IRC | 17:25 | |
*** bigdogstl has quit IRC | 17:30 | |
*** bigdogstl has joined #openstack-swift | 17:33 | |
*** mingyu has quit IRC | 17:34 | |
*** hoonetorg has joined #openstack-swift | 17:35 | |
*** bigdogstl has quit IRC | 17:38 | |
*** itlinux has quit IRC | 17:40 | |
*** bigdogstl has joined #openstack-swift | 17:43 | |
zaitcev | tdasilva, clayg: Thanks, guys. If we could make it even later, it would be okay, but 18:00Z is fine. | 17:57 |
zaitcev | I mean it could be even better. | 17:57 |
zaitcev | I mostly made the patch presentable for the PTG, dropped most offensive XXX-es and tested it. | 17:57 |
tdasilva | zaitcev: maybe 20:00 UTC? | 17:57 |
zaitcev | tdasilva: even better | 17:58 |
tdasilva | ok, will send invite out | 17:58 |
clayg | I think later starts to keep Alistair up pretty late (8pm) - but he does tend to be flexible | 17:58 |
acoles | what? who? | 17:59 |
zaitcev | yeah, better not then | 17:59 |
*** tesseract has quit IRC | 17:59 | |
acoles | OIC yeah I'm at UTC+1 and cschwede UTC+2 | 17:59 |
*** bigdogstl has quit IRC | 17:59 | |
acoles | but I think I ticked 2000UTC on doodle | 18:00 |
clayg | ^ see how he takes all the blame | 18:00 |
acoles | lol | 18:00 |
acoles | it was late when I ticked it ;) | 18:00 |
zaitcev | My wife has an oral surgery in the morning on 9/26, but she's digging the ground with her hove to drive herself to the surgeon and back, as Russians say. | 18:00 |
clayg | why are we doing it *at all* on Tuesday then!? :P | 18:02 |
zaitcev | I don't want to postpone it forever. At least I want you guys to have a good look at what's going on. Otherwise maybe the whole thing is garbage and I don't even know about it. | 18:03 |
clayg | sure, but like... Thursday :P | 18:05 |
clayg | or Monday! | 18:05 |
clayg | or Tuesday... we could push it back an hour | 18:05 |
zaitcev | So, 19:00Z then? | 18:05 |
*** _ix has joined #openstack-swift | 18:11 | |
*** juzuluag has joined #openstack-swift | 18:18 | |
*** mvk has joined #openstack-swift | 18:33 | |
*** mingyu has joined #openstack-swift | 18:35 | |
*** mingyu has quit IRC | 18:40 | |
*** ChubYann has joined #openstack-swift | 18:49 | |
*** juzuluag has quit IRC | 18:49 | |
*** pcaruana has quit IRC | 18:56 | |
*** bigdogstl has joined #openstack-swift | 19:00 | |
*** bigdogstl has quit IRC | 19:07 | |
openstackgerrit | Tim Burke proposed openstack/swift feature/deep: Address some review comments https://review.openstack.org/506759 | 19:31 |
*** Sukhdev has joined #openstack-swift | 19:35 | |
*** Sukhdev has quit IRC | 19:40 | |
openstackgerrit | Merged openstack/swift master: Follow-up test fixup https://review.openstack.org/505454 | 19:41 |
*** silor has quit IRC | 19:47 | |
*** SkyRocknRoll has joined #openstack-swift | 20:09 | |
tdasilva | zaitcev, clayg: I picked those Tuesday times really based on trying to maximize highest number of people based on the doodle table, but we can be flexible... | 20:31 |
clayg | zaitcev: can you "fix" your availability then so that tdasilva can pick times for when people are acctually willing to be there? | 20:33 |
tdasilva | yeah, maybe we can move for later in the week so you don't have to worry about Tuesday and your wife surgery | 20:34 |
zaitcev | Any other day is okay. | 20:38 |
tdasilva | maybe Wed at 1400UTC? | 20:38 |
zaitcev | 14-5=9 | 20:38 |
zaitcev | good enough | 20:38 |
tdasilva | sorry, wrong time...messing up UTC conversion | 20:39 |
zaitcev | although it's 8 in CA, isn't it | 20:39 |
zaitcev | no, even 7 a.m. | 20:39 |
tdasilva | I meant 1600UTC on Wed | 20:39 |
clayg | I'm *technically* awake at 7am - but you may not like me | 20:40 |
tdasilva | lol | 20:40 |
clayg | ... err you might dislike me *more* w/e | 20:40 |
zaitcev | 11 a.m. Central works great for me. | 20:40 |
*** bigdogstl has joined #openstack-swift | 20:42 | |
tdasilva | ok, will update invite | 20:42 |
*** SkyRocknRoll_ has joined #openstack-swift | 20:46 | |
*** bigdogstl has quit IRC | 20:47 | |
tdasilva | ok, in case others want to participate in the meeting, we can share link here: https://bluejeans.com/704762476 and I can also forward calendar invite | 20:48 |
clayg | ... and the new official time is? | 20:49 |
tdasilva | mattoliverau, kota_, rledisez, ^^^ | 20:49 |
tdasilva | Wednesday (09/27) at 1600UTC | 20:49 |
clayg | tdasilva: thanks | 20:49 |
*** SkyRocknRoll_ has quit IRC | 20:51 | |
clayg | well but... that's bug triage day! maybe it'll be our kickoff. | 20:52 |
*** mtreinish_ has joined #openstack-swift | 21:01 | |
*** mtreinish has quit IRC | 21:03 | |
*** mtreinish_ is now known as mtreinish | 21:03 | |
*** SkyRocknRoll_ has joined #openstack-swift | 21:04 | |
*** SkyRocknRoll_ has quit IRC | 21:26 | |
*** hoonetorg has quit IRC | 21:30 | |
*** SkyRocknRoll has quit IRC | 21:32 | |
*** vint_bra has quit IRC | 21:50 | |
*** chlong has quit IRC | 21:51 | |
*** bigdogstl has joined #openstack-swift | 21:53 | |
*** rcernin has quit IRC | 21:54 | |
*** bigdogstl has quit IRC | 21:58 | |
notmyname | how did I just discover this? I'm sure you all knew about it all along http://status.openstack.org/openstack-health/#/g/project/openstack~2Fswift | 22:10 |
timburke | it's not super useful though -- not nearly all of the jobs are represented | 22:13 |
notmyname | yeah. seems to change based on the history selected | 22:14 |
timburke | poking around that api though, you can at least get some things like job timings out of it -- see http://health.openstack.org/build_name/gate-tempest-dsvm-neutron-full-ubuntu-xenial/runs for example | 22:17 |
openstackgerrit | Merged openstack/swift feature/deep: sharding: make find_shard_ranges a broker method https://review.openstack.org/505763 | 22:30 |
*** catintheroof has joined #openstack-swift | 23:07 | |
*** Sukhdev has joined #openstack-swift | 23:26 | |
*** chsc has quit IRC | 23:38 | |
*** _ix has quit IRC | 23:44 | |
*** catintheroof has quit IRC | 23:55 | |
*** alenavolk has joined #openstack-swift | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!