| *** rmcall has quit IRC | 00:08 | |
| *** kyles_ne has quit IRC | 00:12 | |
| *** kyles_ne has joined #openstack-swift | 00:12 | |
| *** RockKuo_Office has joined #openstack-swift | 00:13 | |
| *** kyles_ne has quit IRC | 00:16 | |
| *** gyee has quit IRC | 00:18 | |
| openstackgerrit | A change was merged to openstack/swift: account to account copy implementation https://review.openstack.org/72157 | 00:22 |
|---|---|---|
| *** dmorita has joined #openstack-swift | 00:24 | |
| *** tdasilva has quit IRC | 00:25 | |
| *** judd7 has quit IRC | 00:31 | |
| openstackgerrit | Samuel Merritt proposed a change to openstack/swift: Move multipart MIME parser into utils https://review.openstack.org/120652 | 00:31 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-swiftclient: Updated from global requirements https://review.openstack.org/89250 | 00:31 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 00:31 |
| *** acoles has quit IRC | 00:32 | |
| *** judd7 has joined #openstack-swift | 00:32 | |
| *** judd7 has quit IRC | 00:36 | |
| *** judd7 has joined #openstack-swift | 00:37 | |
| *** mwstorer has quit IRC | 00:38 | |
| *** tdasilva has joined #openstack-swift | 00:40 | |
| *** IgnacioCorderi has quit IRC | 00:53 | |
| *** judd7 has quit IRC | 00:54 | |
| *** judd7 has joined #openstack-swift | 00:56 | |
| *** annegentle has joined #openstack-swift | 01:15 | |
| *** IgnacioCorderi has joined #openstack-swift | 01:16 | |
| *** annegentle has quit IRC | 01:26 | |
| *** judd7 has quit IRC | 01:28 | |
| *** judd7 has joined #openstack-swift | 01:29 | |
| *** annegentle has joined #openstack-swift | 01:29 | |
| *** tkay_ has joined #openstack-swift | 01:30 | |
| *** judd7 has quit IRC | 01:32 | |
| *** tkay has quit IRC | 01:33 | |
| *** tkay_ is now known as tkay | 01:33 | |
| *** judd7 has joined #openstack-swift | 01:34 | |
| *** judd7 has quit IRC | 01:34 | |
| *** judd7 has joined #openstack-swift | 01:35 | |
| *** oomichi has joined #openstack-swift | 01:38 | |
| *** judd7 has quit IRC | 01:39 | |
| *** judd7 has joined #openstack-swift | 01:41 | |
| *** judd7 has quit IRC | 01:44 | |
| *** annegentle has quit IRC | 01:45 | |
| *** judd7 has joined #openstack-swift | 01:45 | |
| *** judd7 has quit IRC | 01:46 | |
| *** nosnos has joined #openstack-swift | 01:47 | |
| *** judd7 has joined #openstack-swift | 01:47 | |
| *** haomaiwang has quit IRC | 01:54 | |
| *** haomaiwang has joined #openstack-swift | 01:54 | |
| *** tkay has quit IRC | 01:57 | |
| *** tdasilva has quit IRC | 02:04 | |
| *** haomaiwang has quit IRC | 02:12 | |
| *** haomaiwang has joined #openstack-swift | 02:13 | |
| *** annegentle has joined #openstack-swift | 02:15 | |
| *** zaitcev has quit IRC | 02:17 | |
| *** annegentle has quit IRC | 02:21 | |
| openstackgerrit | Matthew Oliver proposed a change to openstack/swift: Add concurrent reads option to proxy https://review.openstack.org/117710 | 02:26 |
| *** annegentle has joined #openstack-swift | 02:32 | |
| *** IgnacioCorderi has quit IRC | 02:33 | |
| *** sileht has quit IRC | 02:46 | |
| *** dmsimard_away is now known as dmsimard | 02:49 | |
| *** dmsimard is now known as dmsimard_away | 02:52 | |
| *** vu has joined #openstack-swift | 03:08 | |
| *** vu has quit IRC | 03:19 | |
| *** vu has joined #openstack-swift | 03:19 | |
| *** mrsnivvel has joined #openstack-swift | 03:20 | |
| *** annegentle has quit IRC | 03:21 | |
| *** thurloat has quit IRC | 03:27 | |
| *** thurloat has joined #openstack-swift | 03:27 | |
| *** RockKuo_Office has quit IRC | 03:43 | |
| *** RockKuo_Office has joined #openstack-swift | 03:46 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add support for multiple container-reconciler https://review.openstack.org/103779 | 03:51 |
| *** haomaiwang has quit IRC | 03:54 | |
| *** haomaiwa_ has joined #openstack-swift | 03:54 | |
| *** haomai___ has joined #openstack-swift | 04:00 | |
| *** haomai___ has quit IRC | 04:00 | |
| *** haomaiw__ has joined #openstack-swift | 04:01 | |
| *** haomaiwa_ has quit IRC | 04:02 | |
| *** haomaiwang has joined #openstack-swift | 04:06 | |
| *** haomaiw__ has quit IRC | 04:10 | |
| *** annegentle has joined #openstack-swift | 04:11 | |
| *** echevemaster has quit IRC | 04:22 | |
| *** IgnacioCorderi has joined #openstack-swift | 04:29 | |
| *** HenryG is now known as HenryG_zzzz | 04:40 | |
| *** annegentle has quit IRC | 04:47 | |
| *** ppai has joined #openstack-swift | 04:49 | |
| *** IgnacioCorderi has quit IRC | 04:58 | |
| *** RockKuo_Office has quit IRC | 05:05 | |
| *** IgnacioCorderi has joined #openstack-swift | 05:22 | |
| *** RockKuo_Office has joined #openstack-swift | 05:22 | |
| *** judd7 has quit IRC | 05:23 | |
| *** IgnacioCorderi has quit IRC | 05:29 | |
| *** thurloat has quit IRC | 05:30 | |
| *** thurloat has joined #openstack-swift | 05:30 | |
| *** kopparam has joined #openstack-swift | 05:30 | |
| *** bvandenh has joined #openstack-swift | 05:41 | |
| *** nshaikh has joined #openstack-swift | 05:53 | |
| *** RockKuo_Office has quit IRC | 05:57 | |
| *** IgnacioCorderi has joined #openstack-swift | 06:00 | |
| *** oomichi_ has joined #openstack-swift | 06:03 | |
| *** oomichi has quit IRC | 06:05 | |
| *** navid__ has joined #openstack-swift | 06:10 | |
| *** IgnacioCorderi has quit IRC | 06:11 | |
| *** nshaikh has quit IRC | 06:11 | |
| openstackgerrit | Gil Vernik proposed a change to openstack/swift: On demand data migration for Swift https://review.openstack.org/64430 | 06:12 |
| *** RockKuo_Office has joined #openstack-swift | 06:13 | |
| ahale | oh concurrent proxy read! thats awesome mattoliverau :) | 06:28 |
| *** k4n0 has joined #openstack-swift | 06:34 | |
| *** IgnacioCorderi has joined #openstack-swift | 06:54 | |
| *** sileht has joined #openstack-swift | 07:03 | |
| *** navid__ is now known as nshaikh | 07:03 | |
| *** haomaiwang has quit IRC | 07:10 | |
| *** haomaiwa_ has joined #openstack-swift | 07:10 | |
| *** kopparam has quit IRC | 07:11 | |
| *** kopparam has joined #openstack-swift | 07:12 | |
| openstackgerrit | Matthew Oliver proposed a change to openstack/swift: Add container reverse listings and basic ordering https://review.openstack.org/120709 | 07:16 |
| *** kopparam has quit IRC | 07:17 | |
| *** haomaiwa_ has quit IRC | 07:17 | |
| *** haomaiw__ has joined #openstack-swift | 07:17 | |
| mattoliverau | ahale: thanks :) | 07:21 |
| mattoliverau | just uploaded an attempt for reverse listings, so on that note I'm calling it a night, it's been a long day :) | 07:22 |
| ahale | nice | 07:23 |
| ahale | yeah just yesterday I was muttering how nice it would be to have parallel proxy reads - after looking at a really horrible req that hit two recoverable_node_timeouts before finally getting a response | 07:24 |
| *** sileht has quit IRC | 07:29 | |
| *** sileht has joined #openstack-swift | 07:29 | |
| *** haomaiwang has joined #openstack-swift | 07:31 | |
| mattoliverau | Well I obviously I read your mind ;) | 07:34 |
| *** haomaiw__ has quit IRC | 07:35 | |
| hugokuo | magic .... | 07:40 |
| *** kopparam has joined #openstack-swift | 07:41 | |
| *** IgnacioCorderi has quit IRC | 07:42 | |
| *** oomichi_ has quit IRC | 07:45 | |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Return correct number of changed partitions https://review.openstack.org/120713 | 07:46 |
| *** kopparam has quit IRC | 08:00 | |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Return correct number of changed partitions https://review.openstack.org/120713 | 08:02 |
| omame | nice indeed. hope it'll land soon | 08:06 |
| *** geaaru has joined #openstack-swift | 08:10 | |
| *** oomichi has joined #openstack-swift | 08:52 | |
| *** mkollaro has joined #openstack-swift | 08:54 | |
| *** vr2 has joined #openstack-swift | 08:59 | |
| *** vr2 has quit IRC | 09:03 | |
| *** vr3 has joined #openstack-swift | 09:03 | |
| *** oomichi has quit IRC | 09:10 | |
| *** mkollaro has quit IRC | 09:14 | |
| *** mkollaro has joined #openstack-swift | 09:26 | |
| *** aix has joined #openstack-swift | 09:26 | |
| *** acoles_ has joined #openstack-swift | 09:34 | |
| *** ChanServ sets mode: +v acoles_ | 09:34 | |
| openstackgerrit | Alistair Coles proposed a change to openstack/swift: Fix broken www-authenticate functional test https://review.openstack.org/120737 | 09:37 |
| acoles_ | cschwede_: ^^ fix for gate failures, if you could review | 09:39 |
| openstackgerrit | Alistair Coles proposed a change to openstack/swift: Fix broken www-authenticate functional test https://review.openstack.org/120737 | 09:40 |
| *** haomaiwang has quit IRC | 09:42 | |
| *** haomaiwang has joined #openstack-swift | 09:42 | |
| *** nshaikh has quit IRC | 09:51 | |
| *** DisneyRicky has quit IRC | 09:51 | |
| *** nshaikh has joined #openstack-swift | 09:51 | |
| *** haomaiwang has quit IRC | 10:02 | |
| *** haomaiwang has joined #openstack-swift | 10:03 | |
| *** d0ugal has quit IRC | 10:08 | |
| *** d0ugal has joined #openstack-swift | 10:08 | |
| *** DisneyRicky has joined #openstack-swift | 10:14 | |
| openstackgerrit | Alistair Coles proposed a change to openstack/swift: Add description option to policy config https://review.openstack.org/120424 | 10:18 |
| acoles_ | peluse: when you're awake, could you take a look at https://review.openstack.org/120737 - looks like all swift jobs are failing at the mo | 11:16 |
| acoles_ | and we have >24 hour lag to merge :( | 11:16 |
| acoles_ | portante: ^^ ditto, if you are able | 11:25 |
| *** RockKuo_Office has quit IRC | 11:52 | |
| openstackgerrit | Donagh McCabe proposed a change to openstack/swift: Add swift-checker utility for use by monitoring tools https://review.openstack.org/99961 | 11:56 |
| *** dmorita has quit IRC | 12:11 | |
| *** geaaru has quit IRC | 12:14 | |
| *** ppai has quit IRC | 12:17 | |
| *** nosnos has quit IRC | 12:22 | |
| *** aswadr has joined #openstack-swift | 12:22 | |
| *** annegentle has joined #openstack-swift | 12:22 | |
| *** nosnos has joined #openstack-swift | 12:22 | |
| *** geaaru has joined #openstack-swift | 12:27 | |
| *** nosnos has quit IRC | 12:27 | |
| *** annegentle has quit IRC | 12:28 | |
| *** cschwede has joined #openstack-swift | 12:36 | |
| *** cschwede_ has quit IRC | 12:40 | |
| *** cschwede_ has joined #openstack-swift | 12:40 | |
| *** cschwede_ has quit IRC | 12:41 | |
| *** cschwede has quit IRC | 12:43 | |
| *** cschwede has joined #openstack-swift | 12:44 | |
| *** cschwede has quit IRC | 12:44 | |
| *** cschwede has joined #openstack-swift | 12:46 | |
| *** HenryG_zzzz is now known as HenryG | 12:55 | |
| *** miqui has joined #openstack-swift | 12:55 | |
| *** achhabr has joined #openstack-swift | 13:04 | |
| *** dolphm has joined #openstack-swift | 13:13 | |
| *** annegentle has joined #openstack-swift | 13:23 | |
| *** dmsimard_away is now known as dmsimard | 13:26 | |
| dolphm | is alistair coles here? (acoles_ ?) | 13:27 |
| *** annegentle has quit IRC | 13:28 | |
| dolphm | {check|gate}-swift-dsvm-functional appears to be completely broken due to bug 1368048 | 13:32 |
| dolphm | alistair posted https://review.openstack.org/#/c/120737/ as the fix, but it appears to be failing swiftstack cluster ci fairly consistently | 13:32 |
| dolphm | keystone, swift, and openstack/requirements are all unable to merge anything without a fix -- if anyone has time to help land a solution, please ping me! thanks | 13:34 |
| joearnold | dolphm: early in California at the moment. | 13:34 |
| dolphm | joearnold: i realize. it's early in texas too :) | 13:34 |
| joearnold | dolphm: when the guys/gals get in, they'll get the playback. | 13:34 |
| joearnold | dolphm: True! | 13:35 |
| dolphm | acoles_ was looking into it at ~3am my time -- anyone know what timezone he's in? :-/ | 13:35 |
| peluse | dolphm, just approved 120737 | 13:40 |
| dolphm | peluse: thank you! but is the swiftstack ci failure a concern? | 13:40 |
| dolphm | i'm not familiar with that job | 13:41 |
| dolphm | peluse: if not, i'll ask that the change be rushed to the front of the gate | 13:41 |
| acoles_ | dolphm: just been afk (i'm in UK) - the swiftstack CI test is new to me and I can't figure why its failing | 13:41 |
| peluse | dolphm, that would be notmyname's call but I would say no. its a new test cluster just recently brought online | 13:42 |
| acoles_ | the failed tests seem completely unrelated | 13:42 |
| peluse | yup | 13:42 |
| dolphm | acoles_: thank you for the quick turnaround on that patch. i came into the office to find a bunch of failed gate jobs. nice to have an analysis and fix already out there :) | 13:42 |
| dolphm | acoles_: ack | 13:42 |
| dolphm | i'll ask for the fix to be rushed then | 13:42 |
| acoles_ | dolphm: was kind of fortunate serendipity that I'd noticed the middleware patch a while back | 13:43 |
| acoles_ | dolphm: be great if you can get it bumped up the gate queue | 13:44 |
| dolphm | acoles_: fungi is looking at it now | 13:45 |
| acoles_ | notmyname: torgomatic: clayg: any of you guys have insight on the swiftstack CI test failures here https://review.openstack.org/#/c/120737/ ? | 13:47 |
| acoles_ | there's a bunch of KeyErrors on test config missing 'account2', I looked in the test.conf and there are no account options, tried similar conf on my SAIO and sure enough those test fail, so how do they ever pass? | 13:48 |
| openstackgerrit | paul luse proposed a change to openstack/swift: Early framework for the EC Reconstructor https://review.openstack.org/106910 | 13:50 |
| acoles_ | peluse: ^^ that gate bug set me back looking at this but I am intending to get to it ... sooon | 13:52 |
| peluse | acoles_, no sweat. Keep in mind that its incomplete on purpose, want to land the framework to build upon on the integration rbanch to avoid a ton of dependency chains | 13:52 |
| acoles_ | peluse: sure | 13:53 |
| dolphm | acoles_: peluse: fix is at the front of the gate, btw. ETA 31 min. | 14:01 |
| acoles_ | dolphm: great, thanks | 14:01 |
| peluse | cool | 14:02 |
| *** tongli has joined #openstack-swift | 14:08 | |
| *** jergerber has joined #openstack-swift | 14:18 | |
| *** annegentle has joined #openstack-swift | 14:21 | |
| *** annegentle has quit IRC | 14:21 | |
| *** tdasilva has joined #openstack-swift | 14:22 | |
| *** annegentle has joined #openstack-swift | 14:25 | |
| openstackgerrit | paul luse proposed a change to openstack/swift: Add container_count to policy_stat table https://review.openstack.org/120795 | 14:36 |
| *** mahatic has joined #openstack-swift | 14:45 | |
| *** smart_developer1 has quit IRC | 14:58 | |
| *** smart_developer has joined #openstack-swift | 14:58 | |
| openstackgerrit | A change was merged to openstack/swift: Fix broken www-authenticate functional test https://review.openstack.org/120737 | 15:00 |
| *** k4n0 has quit IRC | 15:01 | |
| *** nshaikh has quit IRC | 15:08 | |
| *** dolphm has left #openstack-swift | 15:11 | |
| acoles_ | ^^ gate should be fixed now | 15:16 |
| *** rmcall has joined #openstack-swift | 15:19 | |
| *** gyee has joined #openstack-swift | 15:37 | |
| *** mahatic has quit IRC | 15:38 | |
| *** bvandenh has quit IRC | 15:46 | |
| *** kenhui has joined #openstack-swift | 15:48 | |
| peluse | acoles_, nice work! | 15:49 |
| *** mahatic has joined #openstack-swift | 15:52 | |
| *** bkopilov has quit IRC | 15:58 | |
| *** aswadr has quit IRC | 16:03 | |
| *** kyles_ne has joined #openstack-swift | 16:08 | |
| *** bkopilov has joined #openstack-swift | 16:12 | |
| *** bkopilov has quit IRC | 16:20 | |
| *** annegentle has quit IRC | 16:22 | |
| *** annegentle has joined #openstack-swift | 16:23 | |
| *** mahatic has quit IRC | 16:24 | |
| *** mkollaro has quit IRC | 16:26 | |
| *** mahatic has joined #openstack-swift | 16:27 | |
| *** vr3 has quit IRC | 16:32 | |
| *** mahatic has quit IRC | 16:33 | |
| *** tkay has joined #openstack-swift | 16:41 | |
| notmyname | good morning | 16:42 |
| notmyname | looks like I missed some excitement | 16:42 |
| notmyname | acoles_: cschwede: thanks for patching reviewing | 16:42 |
| notmyname | acoles_: what's the summary of it? | 16:42 |
| notmyname | seems that keystone started adding their own www-authenticate header? | 16:43 |
| *** kenhui has quit IRC | 16:44 | |
| acoles_ | notmyname: yup. summary here https://bugs.launchpad.net/swift/+bug/1368048/comments/6 | 16:44 |
| notmyname | acoles_: ok, and the reason it passes now is because we now have a list of values for that header. and the test now just tests that our expected value is in there somewhere instead of being the only value | 16:45 |
| acoles_ | notmyname: story starts back here https://bugs.launchpad.net/swift/+bug/1349364 | 16:45 |
| notmyname | correct? | 16:45 |
| *** r-daneel has joined #openstack-swift | 16:45 | |
| acoles_ | notmyname: correct. | 16:45 |
| notmyname | ok, great. thanks for taking care of it | 16:46 |
| acoles_ | pragmatic quick fix | 16:46 |
| dmsimard | A PUT to create a container without a content-length header works on a swift-proxy but it looks like nginx doesn't like that too much. My understanding it is expected behavior - if the body is zero length, the headers should provide Content-Length: 0 (which python-swiftclient does, for instance) | 16:46 |
| notmyname | do you expect any follow-up patches to make it better or is it fine now? | 16:46 |
| acoles_ | notmyname: did you see my query about the swiftstack CI failures (in scrollback) | 16:46 |
| dmsimard | So who should I blame ? swift-proxy for accepting something that shouldn't work (non-standard) or nginx ? :) | 16:47 |
| notmyname | dmsimard: I'd want to look at the spec. eg I know that swift rejects object writes without a content-length that aren't chunked transfer encoding | 16:48 |
| *** mahatic has joined #openstack-swift | 16:48 | |
| acoles_ | notmyname: ie. the swifstack CI failed my patch but we took an executive decision to ignore it. I spent some time trying to figure out the failures but could not make sense of the tests pass with the test.conf being used | 16:48 |
| openstackgerrit | A change was merged to openstack/swift: make the bind_port config setting required https://review.openstack.org/118200 | 16:48 |
| notmyname | dmsimard: unrelated to your question though, I recommend that you not use nginx in front of your swift cluster | 16:48 |
| dmsimard | notmyname: Right. I'm speaking strictly about container creation - I already have the chunked transfer without content-length covered | 16:48 |
| notmyname | acoles_: do you expect any follow-up patches to make the www-authenticate better or is it fine now? | 16:48 |
| dmsimard | notmyname: I'm in fact running tengine, not nginx. tengine handles SSL termination and does not buffer body content. | 16:49 |
| notmyname | dmsimard: :-) | 16:49 |
| acoles_ | notmyname: i think its fine, there's a nit in my patch that cschwede pointed out that the perfectionist in me wants to fix | 16:49 |
| openstackgerrit | A change was merged to openstack/swift: test tempurl header sanitization priority https://review.openstack.org/119885 | 16:50 |
| notmyname | dmsimard: also, I'm not sure we could easily change that behavior in swift without breaking existing clients | 16:50 |
| acoles_ | notmyname: we could check the other www-authenticate header but we'd just be making out test brittle w.r.t. keystonemiddleware changing again | 16:50 |
| notmyname | acoles_: ok, great. last question on that (before moving to swiftstack CI): is there anything that we need to do that could have prevented this from happening, or is this simply the result of good practices and we caught it as we should have? | 16:51 |
| acoles_ | s/out/our/ | 16:51 |
| notmyname | acoles_: ok. I think it's reasonable to not tie our functional tests to a particular auth system | 16:51 |
| dmsimard | notmyname: Yeah, that's my point. Right now the only solution I see is to have this third party library update their API call | 16:51 |
| acoles_ | notmyname: so it reveals that keystonemiddleware isn't gated on swift func tests, i don't know if its reasonable to expect that, i was kinda surprised that no tests flagged the problem before the middleware change landed. | 16:55 |
| *** aix has quit IRC | 16:55 | |
| acoles_ | actually, scrub that last phrase, not surprised - nothing functional got broken, just the swift test | 16:57 |
| notmyname | acoles_: ok, thanks. good to know. I'm perfectly happy saying that the gate issue is the symptom of things working instead of things breaking | 16:58 |
| notmyname | acoles_: ok, now for the swiftstack CI stuff | 16:58 |
| notmyname | acoles_: I also saw some playback from charz in a different channel. the issue is that it is/was running swiftstack auth instead of tempauth | 16:59 |
| smart_developer | Hi, I am trying to run cosbench on my Swift cluster (which has HAProxy as a load balancer, and Keystone for authentication). When configuring a workload with COSBench, does anyone know whether I need to specify both the 'Authentication' and 'Storage' fields? | 17:00 |
| smart_developer | (or rather, the 'Configuration' portions of 'Authentication' and 'Storage') | 17:00 |
| notmyname | acoles_: and some recent patch added an ACL test without a SkipTest. so since the swiftstack auth wasn't configured to use the accounts in a way to do the acls right, the test failed | 17:01 |
| *** mahatic has quit IRC | 17:01 | |
| notmyname | acoles_: and the longer-term fix(es) is to add a SkipTest in swift and add functionality to swiftstack-auth, but the short term fix is to start running with tempauth | 17:02 |
| smart_developer | Basically, I have my cluster set up so that testing with Swift client works when pointing to the HAProxy (which is desired). | 17:02 |
| peluse | smart_developer, there are example configs in the PDF, let me see if I can find one that I use | 17:02 |
| acoles_ | notmyname: yeah, i noticed swiftauth in the logs (pipeline listing) - but when I changed my test.conf to look like the swiftstack CI one, and ran against tempauth, I saw same errors (from a bunch of SLO tests) | 17:03 |
| acoles_ | notmyname: and weird that concurrent jobs to mine passed ok | 17:03 |
| notmyname | acoles_: do you still have that test.conf handy? can you pastebinit and I'll try it too | 17:06 |
| peluse | smart_developer, so I'm not using keystone but I looked in the user guide and it has a clear example. For storage all you need is <storage type="swift"/> | 17:06 |
| acoles_ | notmyname: i'll recreate it | 17:07 |
| peluse | smart_developer, and you might get quicker response for cobench Q&A at https://github.com/intel-cloud/cosbench | 17:07 |
| peluse | smart_developer, for the <storage> section you only need to specicy a config if you want to use a specific swift storage policy, like this: <storage type="swift" config="policy=some-policy"/> | 17:09 |
| *** mahatic has joined #openstack-swift | 17:09 | |
| dmsimard | notmyname: What about deprecating the behavior with a missing content-length ? Otherwise we'd never be able to get rid of it without breaking things. | 17:10 |
| *** IgnacioCorderi has joined #openstack-swift | 17:10 | |
| *** IgnacioCorderi has quit IRC | 17:11 | |
| notmyname | dmsimard: maybe, but since we have clients that were written 4 or 5 years ago still talking to swift clusters today, it makes it tricky | 17:11 |
| *** pberis has joined #openstack-swift | 17:11 | |
| notmyname | dmsimard: probably it's one more thing on the "well, if we ever bump the API version, this is something to change" | 17:11 |
| tdasilva | dmsimard, notmyname: haven't been following too closely, but are you sure that the proxy checks content-length on container put? it checks on a object PUT, but does it also check on a container PUT? | 17:16 |
| tdasilva | sorry if i'm missing something... | 17:16 |
| notmyname | tdasilva: don't know for sure. I'm still going through IRC buffers and emails from ovenight :-) | 17:16 |
| *** bkopilov has joined #openstack-swift | 17:17 | |
| acoles_ | notmyname: test.conf here http://paste.openstack.org/show/110240/ | 17:17 |
| notmyname | acoles_: thanks | 17:18 |
| acoles_ | test run here with tempauth http://paste.openstack.org/show/110241/ | 17:18 |
| acoles_ | notmyname: | 17:18 |
| acoles_ | notmyname: ^^ | 17:18 |
| notmyname | oh! account to account copy landed! | 17:19 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 17:21 |
| acoles_ | notmyname: found it https://review.openstack.org/#/c/72157/19/test/functional/tests.py | 17:21 |
| *** IgnacioCorderi has joined #openstack-swift | 17:22 | |
| notmyname | acoles_: ah, ok | 17:23 |
| acoles_ | notmyname: in the test setUp line 799 its trying to get config['account2'] which will fail with the 'shorthand' test2:tester2 config | 17:26 |
| *** zaitcev has joined #openstack-swift | 17:27 | |
| *** ChanServ sets mode: +v zaitcev | 17:27 | |
| acoles_ | at first glance it should probably be using swift_test_user[1] | 17:28 |
| notmyname | acoles_: yup. that's sounds exactly like the symptoms charz told me | 17:35 |
| acoles_ | notmyname: i just tried a very quick fix but the test then failed (rather than errored), and I'm out of time for today | 17:37 |
| notmyname | acoles_: ok | 17:37 |
| * acoles_ must finish reviewing EC patch | 17:37 | |
| acoles_ | notmyname: so how does the swiftstack CI job get associated with a review? I don't see it on the account to account copy? | 17:39 |
| notmyname | acoles_: it pulls the feed from gerrit into jenkins and then runs the tests. it uses the gerrit api to post the review comment. and it has an account with -infra so that it gets registered as a test thingy | 17:40 |
| clayg | what's the usecase for reverse and byte-order listings? | 17:42 |
| acoles_ | notmyname: cool. so i'm guessing its triggered by new reviews and not yet running against older reviews? | 17:42 |
| notmyname | clayg: nothing urgent AFAIK. more of something that has been talked about for a while and a "completion" of functionality | 17:44 |
| notmyname | clayg: I could make up a use case, but its not something that has a specific customer request from anyone AFAIK | 17:44 |
| clayg | hrmmmm.... | 17:45 |
| notmyname | clayg: my use case would be "find the stuff at the end of the listing without having to go through from the start" | 17:45 |
| clayg | notmyname: hrmmm... like if the names are dates | 17:46 |
| *** openstackgerrit has quit IRC | 17:46 | |
| clayg | notmyname: you could check marker=2014; marker=2013; but if nothings been added to a container in 50 years it could be borning | 17:47 |
| *** openstackgerrit has joined #openstack-swift | 17:48 | |
| notmyname | acoles_: (assuming I did it correctly) I just set the swiftstack CI to silent mode so it won't pollute the results with false failures | 17:49 |
| acoles_ | notmyname: you couldn't do the same for the other gate jobs could you ;) the keystone ACL patch got burnt by the www-authenticate bug after 26 hours, and has got hit again by a 'random' failure | 17:50 |
| notmyname | ugh | 17:51 |
| *** portante has quit IRC | 17:51 | |
| notmyname | ah, looks like the -qa team is talking about the tempest patch for swift (response code classes) in #openstack-meeting right now | 17:53 |
| *** portante has joined #openstack-swift | 17:55 | |
| *** ChanServ sets mode: +v portante | 17:55 | |
| *** rmcall has quit IRC | 17:57 | |
| *** mahatic has quit IRC | 17:58 | |
| notmyname | looks like they're plannign on merging it as-is | 17:59 |
| *** rmcall has joined #openstack-swift | 18:01 | |
| *** kenhui has joined #openstack-swift | 18:04 | |
| torgomatic | clayg: also the object versioning code could make use of reverse listings; right now it lists the whole container just to get the [-1] element | 18:05 |
| portante | torgomatic: like all 10,000+ objects? | 18:05 |
| torgomatic | portante: yep, that's the potentially-unbounded one | 18:06 |
| notmyname | torgomatic: oh yeah, I remember that use case now | 18:06 |
| torgomatic | swift/proxy/controllers/obj.py line 800-ish | 18:06 |
| *** geaaru has quit IRC | 18:07 | |
| zaitcev | well, if sqlite can do reversing efficiently why the heck not | 18:08 |
| zaitcev | reading directories in reverse is a pain, but containers are different | 18:09 |
| zaitcev | gluster is gonna cry bloody tears tho | 18:09 |
| notmyname | zaitcev: well except that for now gluster isn't backing container DBs | 18:15 |
| zaitcev | notmyname: oh, right. | 18:15 |
| *** echevemaster has joined #openstack-swift | 18:19 | |
| *** tkay has quit IRC | 18:20 | |
| *** tkay has joined #openstack-swift | 18:23 | |
| *** kenhui has quit IRC | 18:33 | |
| *** IgnacioCorderi has quit IRC | 18:38 | |
| *** nshaikh has joined #openstack-swift | 18:42 | |
| *** acoles_ has quit IRC | 18:52 | |
| *** zacksh has quit IRC | 19:03 | |
| *** StevenK has quit IRC | 19:06 | |
| *** rmcall has quit IRC | 19:08 | |
| *** StevenK_ has joined #openstack-swift | 19:12 | |
| *** StevenK_ has quit IRC | 19:17 | |
| *** StevenK_ has joined #openstack-swift | 19:17 | |
| *** annegentle has quit IRC | 19:18 | |
| *** jroll has quit IRC | 19:19 | |
| *** StevenK_ has quit IRC | 19:22 | |
| *** StevenK_ has joined #openstack-swift | 19:23 | |
| *** vu_ has joined #openstack-swift | 19:23 | |
| *** vu has quit IRC | 19:24 | |
| *** StevenK_ has quit IRC | 19:27 | |
| *** StevenK_ has joined #openstack-swift | 19:28 | |
| *** StevenK_ has quit IRC | 19:33 | |
| *** StevenK_ has joined #openstack-swift | 19:33 | |
| *** smart_developer1 has joined #openstack-swift | 19:37 | |
| *** StevenK_ has quit IRC | 19:38 | |
| *** StevenK_ has joined #openstack-swift | 19:39 | |
| *** jergerber has quit IRC | 19:40 | |
| *** smart_developer has quit IRC | 19:40 | |
| *** nshaikh has left #openstack-swift | 19:43 | |
| *** StevenK_ has quit IRC | 19:44 | |
| *** StevenK_ has joined #openstack-swift | 19:44 | |
| *** Dafna has quit IRC | 19:45 | |
| *** StevenK_ has quit IRC | 19:49 | |
| *** StevenK_ has joined #openstack-swift | 19:50 | |
| *** echevemaster has quit IRC | 19:50 | |
| *** jroll has joined #openstack-swift | 19:51 | |
| *** zacksh has joined #openstack-swift | 19:52 | |
| *** annegentle has joined #openstack-swift | 19:52 | |
| notmyname | questions from sandisk this morning about running swift http://lists.openstack.org/pipermail/openstack-operators/2014-September/005094.html | 20:00 |
| notmyname | portante: you'd probably find that interesting ^^ | 20:01 |
| peluse | hey, they're using the right procs at least :) | 20:02 |
| ahale | i was wishing the greenpool (?) size was configurable the other day, for science | 20:05 |
| *** echevemaster has joined #openstack-swift | 20:06 | |
| notmyname | ahale: like all good scientists, did you end up cutting the subject open and changing the guts to make it do what you wanted? | 20:07 |
| ahale | if that means editting the files manually in prod, yes I did :) | 20:08 |
| notmyname | :-) | 20:08 |
| notmyname | I'm at my desk starting to eat the sriracha-salt fries, and of course now is when my eye starts itching. makes me nervous | 20:09 |
| *** vu_ has quit IRC | 20:10 | |
| *** echevemaster has quit IRC | 20:14 | |
| *** tdasilva has quit IRC | 20:15 | |
| *** IgnacioCorderi has joined #openstack-swift | 20:21 | |
| *** IgnacioCorderi has quit IRC | 20:26 | |
| *** ZBhatti has joined #openstack-swift | 20:40 | |
| peluse | torgomatic, do you have some concerns about dfg's expiring obj deletion patch? | 20:58 |
| torgomatic | peluse: I haven't looked enough to have concerns | 20:58 |
| peluse | torgomatic, ahh, OK. Want me to hold off on the 2nd +2 for ya? | 20:59 |
| torgomatic | but it seemed fairly straightforward... splay out deletes across containers instead of hammering one, then dropping it and hammering another, and so on... | 20:59 |
| peluse | (you +1'd it so wasn't sure) | 21:00 |
| torgomatic | peluse: usually a +1 from me just means I haven't had time to try it out | 21:00 |
| peluse | OK, cool. I'm sending it on to the other side then... | 21:00 |
| torgomatic | but if you think it's good, push the button. no need to wait for me. ;) | 21:00 |
| torgomatic | hooray! | 21:00 |
| openstackgerrit | David Goetz proposed a change to openstack/swift: Add Error Limiting to slow nodes. https://review.openstack.org/112424 | 21:00 |
| peluse | :) | 21:00 |
| smart_developer1 | What's the best way to do some functionality testing on your Swift cluster? | 21:01 |
| smart_developer1 | For instance, to see that all the parts actually work. | 21:02 |
| smart_developer1 | By using either COSBench or SSBench. | 21:02 |
| *** zaitcev has quit IRC | 21:03 | |
| notmyname | wow. HP just bought eucalyptus | 21:05 |
| torgomatic | for reference: http://en.wikipedia.org/wiki/Eucalyptus | 21:07 |
| *** annegentle has quit IRC | 21:09 | |
| *** tongli has quit IRC | 21:12 | |
| *** annegentle has joined #openstack-swift | 21:20 | |
| *** annegentle has quit IRC | 21:23 | |
| *** annegentle has joined #openstack-swift | 21:24 | |
| notmyname | It would seem that DefCore has changed their minds and Swift will be presented to the openstack board of directors as having no core capabilities for openstack clouds | 21:27 |
| notmyname | http://lists.openstack.org/pipermail/defcore-committee/2014-September/000349.html | 21:27 |
| notmyname | this means that in order to call yourself openstack, you do not need to include swift (or swift features) in any way, shape, or form | 21:28 |
| *** IgnacioCorderi has joined #openstack-swift | 21:33 | |
| torgomatic | and HP just bought Eucalyptus too... shall we start a betting pool on how many days until Nova has no designated sections either? | 21:37 |
| swifterdarrell | torgomatic: lol | 21:44 |
| *** StevenK_ is now known as StevenK | 21:47 | |
| peluse | smart_developer1, either of those tools are good for supplying a specfiic workload and measuring performance but neither is intended to specifically test sll of the swift "parts" | 21:52 |
| *** annegentle has quit IRC | 21:54 | |
| *** dmsimard is now known as dmsimard_away | 21:57 | |
| *** vu has joined #openstack-swift | 22:01 | |
| smart_developer1 | by "sll" you meant to say "all", right? | 22:02 |
| smart_developer1 | peluse @ | 22:03 |
| *** vu has quit IRC | 22:03 | |
| peluse | yes, sorry. my keyboard seems to always get things wrong :) | 22:03 |
| notmyname | peluse: you must have the same brand keyboard as me | 22:04 |
| notmyname | peluse: mine has the extra feature of being able to type syntactically correct python that has bugs in it | 22:04 |
| peluse | notmyname, yeah, I keep buying new ones and I always get the defective ones for some reason... | 22:04 |
| *** vu has joined #openstack-swift | 22:04 | |
| *** cschwede has quit IRC | 22:09 | |
| *** achhabr has quit IRC | 22:16 | |
| *** zaitcev has joined #openstack-swift | 22:18 | |
| *** ChanServ sets mode: +v zaitcev | 22:18 | |
| *** ZBhatti_ has joined #openstack-swift | 22:28 | |
| *** ZBhatti has quit IRC | 22:29 | |
| *** ZBhatti_ is now known as ZBhatti | 22:29 | |
| *** cebruns_ has quit IRC | 22:30 | |
| *** cebruns has joined #openstack-swift | 22:31 | |
| *** wer has quit IRC | 22:36 | |
| mattoliverau | Morning | 22:36 |
| *** ZBhatti has quit IRC | 22:38 | |
| *** openstackgerrit has quit IRC | 22:38 | |
| *** openstackgerrit_ has joined #openstack-swift | 22:40 | |
| *** openstackgerrit_ is now known as openstackgerrit | 22:40 | |
| *** gyee has quit IRC | 22:40 | |
| *** wer has joined #openstack-swift | 22:41 | |
| *** AllyUnion has joined #openstack-swift | 22:52 | |
| AllyUnion | Hi, our company recently added 450 X 4 TB hard drives to our 15 machine, 5 zone swift cluster. I've been running the swift-dispersion-report about every two hours, and I've noticed that the pct_found has been slowly decreasing over time. Is this an issue I should be concerned about or expected behavior? | 22:54 |
| *** erlon has quit IRC | 23:00 | |
| notmyname | AllyUnion: did you add the capacity gradually by increasing the weight over time or did you add it all at once? | 23:12 |
| mattoliverau | Yay, it's been over 14 days since the proper notifications have been sent, and there is now list of abandoned changes! | 23:16 |
| notmyname | mattoliverau: cool! | 23:16 |
| notmyname | I've got to head home, but it looks like I have some community stuff to do tonight (gumble*defcore*grumble* | 23:17 |
| notmyname | mattoliverau: I'll try to remember to look at that list tonight | 23:17 |
| mattoliverau | notmyname: thanks, and good luck.... you can always open a new bottle of liquid courage :P | 23:21 |
| zaitcev | If my pct_found were declining after such change, I would suspect that swift-dispersion-populate was somehow missing devices. But I have no comparable operational experience, sorry. | 23:32 |
| *** r-daneel has quit IRC | 23:33 | |
| AllyUnion | notmyname: Gradually by increasing the weight over time (25 at a time per hour) | 23:35 |
| *** oomichi has joined #openstack-swift | 23:37 | |
| *** tkay has quit IRC | 23:37 | |
| *** kyles_ne has quit IRC | 23:37 | |
| *** kyles_ne has joined #openstack-swift | 23:38 | |
| *** kyles_ne has quit IRC | 23:42 | |
| *** tdasilva has joined #openstack-swift | 23:52 | |
| *** aix has joined #openstack-swift | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!