| openstackgerrit | Merged openstack/swift: Document X-Openstack-Request-Id in all responses https://review.openstack.org/397896 | 00:06 |
|---|---|---|
| *** _JZ_ has quit IRC | 00:10 | |
| *** zul has quit IRC | 00:13 | |
| si1ver | Hello again. I'm still trying to get aws4 auth working with swift 3 in mitaka. I started over with a new test node following http://docs.openstack.org/mitaka/install-guide-ubuntu/keystone-install.html, http://docs.openstack.org/mitaka/install-guide-ubuntu/swift-controller-install.html, and http://docs.openstack.org/mitaka/config-reference/object-storage/configure-s3.html. I'm still not able to use aws credentials to access my te | 00:19 |
| si1ver | The instructions above from openstack.org do not mention s3 token, but it is mentioned in the repo readme at https://github.com/openstack/swift3. Is s3token required in addition to swift3? | 00:21 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 00:22 |
| *** sgundur_ has joined #openstack-swift | 00:23 | |
| si1ver | I tried adding s3token to my main pipeline and get this error: LookupError: Entry point 's3token' not found in egg 'swift3'. Does that mean that s3token is not included in the ubuntu cloud archive package swift-plugin-s3? | 00:24 |
| *** kei_yama has quit IRC | 00:24 | |
| openstackgerrit | John Dickinson proposed openstack/swift: Add checksum to object extended attributes https://review.openstack.org/336323 | 00:27 |
| timburke | si1ver: yes, s3token is required when using swift3 with keystone. this middleware used to be provided as part of keystonemiddleware, but was moved into swift3's tree in the not-so-distant past | 00:28 |
| timburke | you might want to look at a slightly older version of the sample configs, like from https://github.com/openstack/swift3/blob/v1.10/etc/proxy-server.conf-sample#L144 | 00:28 |
| si1ver | Ok I removed the swift-plugin-s3 package and ran git clone https://github.com/openstack/swift3.git and the setup script. Truing it now. | 00:28 |
| timburke | or that :-) | 00:28 |
| si1ver | so filter:s3token does not need auth creds in it? The older examples I found had them in there, but https://github.com/openstack/swift3 doesn't show them. | 00:29 |
| si1ver | I have the same error now that I did on my previous test node, but I think I've got a better starting place for debugging now. "The request signature we calculated does not match the signature you provided. Check your key and signing method." | 00:30 |
| timburke | si1ver: nope. i think those were cargo-culted in before. s3token basically takes the client's (slightly massaged & standardized) request, sends it as JSON to keystone, and says "here, you deal with it" | 00:31 |
| timburke | do we see the keystone request in the proxy logs? or for that matter, the keystone logs | 00:31 |
| si1ver | I see this in the proxy log: 'Nov 16 00:30:46 swift3 proxy-server: 10.0.5.94 10.0.5.94 16/Nov/2016/00/30/46 GET /ma-swift%3Fencoding-type%3Durl%26max-keys%3D1 HTTP/1.0 403 - aws-sdk-ruby2/2.6.19%20ruby/2.0.0%20x86_64-darwin14.1.0 - - 277 - tx3e8c4098e9bd454396558-00582ba8b6 - 0.0408 - - 1479256246.784826040 1479256246.825629950 -' | 00:31 |
| si1ver | Looking for the keystone log lines for that request now. | 00:32 |
| si1ver | ma-swift is a container that I created with the normal python swift client as the demo user, which is the same user these ec2 creds were generated by. | 00:32 |
| si1ver | keystone says 'POST http://swift3:35357/v2.0/s3tokens' and 'Authorization failed. The request you have made requires authentication' | 00:33 |
| si1ver | So if I understand correctly, that means that swift itself failed to auth against keystone, so the user creds were never passed along. Does that sound right? | 00:34 |
| timburke | only that keystone request shouldn't require auth :-/ i'm not sure why keystone thinks that it does | 00:35 |
| si1ver | Is s3tokens a new endpoint that I need to create? That's not mentioned in the docs. | 00:35 |
| timburke | possibly. i know the func tests ensure the extension is defined in keystone's paste.ini at https://github.com/openstack/swift3/blob/master/swift3/test/functional/conf/keystone-paste.ini#L36-L37 (and enabled in the pipelines around lines 73 & 76) | 00:38 |
| timburke | we almost certainly could use better docs | 00:38 |
| si1ver | Well I'd love to help. :p | 00:38 |
| si1ver | swift's been good to me | 00:38 |
| openstackgerrit | John Dickinson proposed openstack/swift: Add checksum to object extended attributes https://review.openstack.org/336323 | 00:39 |
| timburke | unfortunately, though, i've gotta run to pick up my daughter | 00:39 |
| si1ver | No sweat. | 00:39 |
| timburke | si1ver: sorry i couldn't stick around to help more | 00:39 |
| *** sgundur_ has quit IRC | 00:43 | |
| *** zul has joined #openstack-swift | 00:43 | |
| *** kei_yama has joined #openstack-swift | 00:58 | |
| *** tqtran has quit IRC | 01:41 | |
| *** juzuluag has joined #openstack-swift | 01:52 | |
| *** links has joined #openstack-swift | 01:55 | |
| *** links has quit IRC | 02:07 | |
| *** siva_krish has joined #openstack-swift | 02:11 | |
| *** winggundamth has quit IRC | 03:02 | |
| si1ver | timburke: I got it working with s3curl, but I am still having trouble with other clients. | 03:04 |
| *** winggundamth has joined #openstack-swift | 03:15 | |
| *** juzuluag has quit IRC | 03:17 | |
| si1ver | Hostfiling s3.amazonaws.com to my testing node and proxying 80 to 8080 doesn't do the trick for the signature in the ruby aws client. Must be more to it, like maybe ssl. | 03:19 |
| *** siva_krish has quit IRC | 03:19 | |
| *** dmorita has quit IRC | 03:22 | |
| *** dmorita has joined #openstack-swift | 03:23 | |
| *** dmorita has quit IRC | 03:27 | |
| *** sanchitmalhotra has quit IRC | 03:46 | |
| *** sanchitmalhotra has joined #openstack-swift | 03:47 | |
| *** psachin has joined #openstack-swift | 03:54 | |
| *** deep_1 has joined #openstack-swift | 03:59 | |
| *** siva_krish has joined #openstack-swift | 04:01 | |
| *** siva_krish has quit IRC | 04:01 | |
| *** Jeffrey4l has joined #openstack-swift | 04:03 | |
| mahatic | good morning | 04:24 |
| *** SkyRocknRoll has joined #openstack-swift | 04:39 | |
| mattoliverau | mahatic: morning | 04:40 |
| mahatic | mattoliverau: o/ | 04:41 |
| openstackgerrit | Mahati Chamarthy proposed openstack/swift: remove empty db hash and suffix directories https://review.openstack.org/331601 | 04:42 |
| *** klrmn has quit IRC | 04:59 | |
| *** darrenc is now known as darrenc_afk | 05:19 | |
| *** darrenc_afk is now known as darrenc | 05:57 | |
| *** Trixboxer has quit IRC | 06:09 | |
| *** ChubYann has quit IRC | 06:10 | |
| *** ppai has joined #openstack-swift | 06:22 | |
| *** sams-gleb has joined #openstack-swift | 07:04 | |
| *** pcaruana has joined #openstack-swift | 07:41 | |
| *** sams-gleb has quit IRC | 07:50 | |
| *** sams-gleb has joined #openstack-swift | 07:51 | |
| *** sams-gleb has quit IRC | 07:53 | |
| *** sams-gleb has joined #openstack-swift | 07:53 | |
| *** sams-gleb has quit IRC | 07:57 | |
| *** rledisez has joined #openstack-swift | 08:13 | |
| *** amoralej|off is now known as amoralej | 08:20 | |
| *** sams-gleb has joined #openstack-swift | 08:21 | |
| *** geaaru has joined #openstack-swift | 08:23 | |
| *** jordanP has joined #openstack-swift | 08:23 | |
| *** dmorita has joined #openstack-swift | 08:23 | |
| *** dmorita has quit IRC | 08:27 | |
| *** jordanP has quit IRC | 08:35 | |
| *** joeljwright has joined #openstack-swift | 08:43 | |
| *** ChanServ sets mode: +v joeljwright | 08:43 | |
| *** joeljwright1 has joined #openstack-swift | 08:57 | |
| *** kei_yama has quit IRC | 08:58 | |
| *** joeljwright2 has joined #openstack-swift | 08:59 | |
| *** joeljwright has quit IRC | 08:59 | |
| *** joeljwright1 has quit IRC | 09:02 | |
| *** jordanP has joined #openstack-swift | 09:23 | |
| *** joeljwright2 has quit IRC | 09:34 | |
| *** joeljwright has joined #openstack-swift | 09:34 | |
| *** ChanServ sets mode: +v joeljwright | 09:34 | |
| *** deep_1 has quit IRC | 09:35 | |
| *** takashi_ has joined #openstack-swift | 09:38 | |
| *** acoles_ is now known as acoles | 09:47 | |
| *** deep_1 has joined #openstack-swift | 09:50 | |
| *** hseipp has joined #openstack-swift | 09:56 | |
| *** dstepanenko_ is now known as dstepanenko | 10:13 | |
| *** asettle has joined #openstack-swift | 10:23 | |
| *** mvk has quit IRC | 10:30 | |
| *** donagh has joined #openstack-swift | 10:43 | |
| *** SkyRocknRoll has quit IRC | 10:43 | |
| *** sams-gle_ has joined #openstack-swift | 10:50 | |
| *** sams-gleb has quit IRC | 10:50 | |
| *** SkyRocknRoll has joined #openstack-swift | 10:52 | |
| *** mvk has joined #openstack-swift | 10:54 | |
| *** geaaru has quit IRC | 10:55 | |
| *** nikivi has joined #openstack-swift | 11:02 | |
| *** sanchitmalhotra has quit IRC | 11:06 | |
| *** geaaru has joined #openstack-swift | 11:09 | |
| *** nikivi has quit IRC | 11:23 | |
| *** deep_1 has quit IRC | 11:31 | |
| *** tqtran has joined #openstack-swift | 11:41 | |
| *** tqtran has quit IRC | 11:45 | |
| *** geaaru has quit IRC | 11:56 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Make tempurl functional tests clean up account keys https://review.openstack.org/346752 | 12:03 |
| *** geaaru has joined #openstack-swift | 12:14 | |
| *** saltsa has quit IRC | 12:14 | |
| *** thurloat has quit IRC | 12:15 | |
| *** saltsa has joined #openstack-swift | 12:16 | |
| *** thurloat has joined #openstack-swift | 12:16 | |
| *** catintheroof has joined #openstack-swift | 12:23 | |
| *** deep_1 has joined #openstack-swift | 12:29 | |
| *** portante has quit IRC | 12:38 | |
| *** portante has joined #openstack-swift | 12:40 | |
| *** nexusz99 has joined #openstack-swift | 12:45 | |
| *** nexusz99 has quit IRC | 12:46 | |
| *** deep_1 has quit IRC | 12:46 | |
| *** takashi_ has quit IRC | 13:10 | |
| *** asettle__ has joined #openstack-swift | 13:15 | |
| *** asettle has quit IRC | 13:18 | |
| *** ppai has quit IRC | 13:19 | |
| *** _JZ_ has joined #openstack-swift | 13:22 | |
| *** jamielennox is now known as jamielennox|away | 13:25 | |
| *** asettle__ is now known as asettle | 13:25 | |
| *** bkopilov_ is now known as bkopilov | 13:28 | |
| *** siva_krish has joined #openstack-swift | 13:31 | |
| *** ouchkernel has quit IRC | 13:39 | |
| *** SkyRocknRoll has quit IRC | 13:39 | |
| *** ndk_ has quit IRC | 13:40 | |
| *** ndk_ has joined #openstack-swift | 13:42 | |
| *** tqtran has joined #openstack-swift | 13:42 | |
| *** siva_krish has left #openstack-swift | 13:44 | |
| *** tqtran has quit IRC | 13:47 | |
| *** Administrator_ has quit IRC | 13:51 | |
| *** Administrator_ has joined #openstack-swift | 13:52 | |
| *** amoralej is now known as amoralej|lunch | 13:52 | |
| *** ouchkernel has joined #openstack-swift | 14:05 | |
| *** deep_1 has joined #openstack-swift | 14:11 | |
| *** klamath has joined #openstack-swift | 14:13 | |
| *** Administrator_ has quit IRC | 14:13 | |
| *** Administrator_ has joined #openstack-swift | 14:14 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Make tempurl functional tests clean up account keys https://review.openstack.org/346752 | 14:16 |
| *** amoralej|lunch is now known as amoralej | 14:32 | |
| tdasilva | good morning | 14:36 |
| *** klamath has quit IRC | 14:40 | |
| *** klamath has joined #openstack-swift | 14:41 | |
| *** SkyRocknRoll has joined #openstack-swift | 14:43 | |
| *** tongli has joined #openstack-swift | 14:44 | |
| *** psachin has quit IRC | 14:51 | |
| *** sams-gle_ has quit IRC | 15:06 | |
| *** sams-gleb has joined #openstack-swift | 15:07 | |
| *** sams-gleb has quit IRC | 15:11 | |
| openstackgerrit | Thiago da Silva proposed openstack/swift: Symlink implementation. https://review.openstack.org/232162 | 15:15 |
| *** dmorita has joined #openstack-swift | 15:24 | |
| *** sams-gleb has joined #openstack-swift | 15:24 | |
| *** dmorita has quit IRC | 15:28 | |
| *** SkyRocknRoll has quit IRC | 15:34 | |
| *** SkyRocknRoll has joined #openstack-swift | 15:38 | |
| *** deep_1 has quit IRC | 15:43 | |
| *** tqtran has joined #openstack-swift | 15:44 | |
| *** tqtran has quit IRC | 15:48 | |
| timburke | good morning | 15:52 |
| tdasilva | timburke: o/ | 15:57 |
| acoles | tdasilva: timburke o/ | 16:02 |
| tdasilva | acoles: hi! you asking me about cross-policy testing in symlinks led me to find a bug in the code :) | 16:03 |
| acoles | tdasilva: great! I forget what I asked but glad it helped :) | 16:04 |
| tdasilva | acoles: back in BCN you just asked if the func tests had cross-policy testing :) | 16:05 |
| acoles | right. because one symlinks is to link across policies. | 16:05 |
| tdasilva | right | 16:06 |
| openstackgerrit | Thiago da Silva proposed openstack/swift: add policy info to swift-get-nodes https://review.openstack.org/372716 | 16:08 |
| openstackgerrit | Thiago da Silva proposed openstack/swift: Fix swiftdir option and usage of storage policy aliases https://review.openstack.org/344693 | 16:08 |
| openstackgerrit | Alexis Lee proposed openstack/swift: fixup https://review.openstack.org/398444 | 16:09 |
| jrichli | tdasilva, acoles: nice | 16:12 |
| acoles | tdasilva: you just helped me realise why my patch is failing in gate. serendipity at work :) | 16:14 |
| timburke | acoles: i was just about to help out there :-) | 16:15 |
| acoles | tdasilva: its failing cross policy versioning tests, which I remembered are skipped in gate | 16:15 |
| tdasilva | acoles: http://logs.openstack.org/52/346752/6/check/gate-swift-tox-func-in-process-fast-post-ubuntu-xenial/d4f873c/console.html | 16:16 |
| tdasilva | that? | 16:16 |
| acoles | yep | 16:16 |
| acoles | caused by calling the teardown when setup doesn't complete cos the tests skip | 16:17 |
| acoles | I don't see that OMM where I have multiple policies | 16:18 |
| tdasilva | oh i see | 16:18 |
| tdasilva | acoles: I was thinking about a patch to start "breaking down" the tests.py | 16:18 |
| tdasilva | into smaller files | 16:18 |
| timburke | i love that idea | 16:18 |
| acoles | tdasilva: pycharm would thank you, it groans a bit with the larger test files | 16:19 |
| *** klrmn has joined #openstack-swift | 16:19 | |
| *** oshritf has quit IRC | 16:19 | |
| *** asettle has quit IRC | 16:20 | |
| *** asettle has joined #openstack-swift | 16:20 | |
| tdasilva | hehehe | 16:20 |
| tdasilva | acoles, timburke I thought one nice thing about test_symlink.py is that I was able to put the skip right on the setupModule | 16:21 |
| tdasilva | skips the whole thing if symlinks is not in the pipeline | 16:22 |
| *** deep_1 has joined #openstack-swift | 16:25 | |
| tdasilva | ntata: hi, I got to test your tempest plugin patch :) | 16:26 |
| tdasilva | ntata: it worked out well, the downside was that I pretty much needed to setup everything I typically do for running functional tests anyway. Meaning download swift repo, install and setup /etc/swift/tests.conf | 16:27 |
| tdasilva | ntata: so the only real change was running tempest instead of .functests | 16:28 |
| *** caiobrentano_ has joined #openstack-swift | 16:36 | |
| *** siva_krish has joined #openstack-swift | 16:43 | |
| *** abalfour has quit IRC | 16:46 | |
| *** abalfour has joined #openstack-swift | 16:46 | |
| *** jistr is now known as jistr|biab | 16:48 | |
| notmyname | good morning | 16:50 |
| notmyname | timburke: thanks for https://review.openstack.org/#/c/398467/ | 16:51 |
| patchbot | patch 398467 - governance - Acknowledge remove-incubated-oslo-code for Swift | 16:51 |
| timburke | np | 16:51 |
| notmyname | I had just started looking at where to make a patch, then I saw the email about your patch. yay | 16:52 |
| openstackgerrit | Alistair Coles proposed openstack/swift: Make tempurl functional tests clean up account keys https://review.openstack.org/346752 | 16:53 |
| acoles | this time be happy jenkins | 16:53 |
| notmyname | nice. looks like I finally got enough tests to skip so that https://review.openstack.org/#/c/336323/ passes the gate | 17:01 |
| patchbot | patch 336323 - swift - Add checksum to object extended attributes | 17:01 |
| notmyname | now to figure out https://review.openstack.org/#/c/394600/ for the infra side so that tests don't get skipped | 17:02 |
| patchbot | patch 394600 - openstack-infra/project-config - enable xfs for swift in-process functests | 17:02 |
| notmyname | if anyone knows anything about zuul or CI, any help on that last patch is appreciated. | 17:02 |
| *** klrmn has quit IRC | 17:09 | |
| *** vint_bra has joined #openstack-swift | 17:15 | |
| *** lxsli is now known as lxsli_away | 17:30 | |
| *** rledisez has quit IRC | 17:31 | |
| *** jistr|biab is now known as jistr | 17:36 | |
| *** sgundur_ has joined #openstack-swift | 17:39 | |
| *** sgundur_ has quit IRC | 17:45 | |
| *** dmorita has joined #openstack-swift | 17:46 | |
| *** david-lyle has quit IRC | 17:48 | |
| *** david-lyle has joined #openstack-swift | 17:48 | |
| *** hseipp has quit IRC | 17:51 | |
| *** SkyRocknRoll has quit IRC | 17:57 | |
| *** mvk has quit IRC | 17:58 | |
| *** cdelatte has joined #openstack-swift | 18:00 | |
| *** deep_1 has quit IRC | 18:02 | |
| *** klrmn has joined #openstack-swift | 18:02 | |
| clayg | timburke: wow, nice work -> https://review.openstack.org/#/c/398467/1/goals/ocata/remove-incubated-oslo-code.rst | 18:04 |
| patchbot | patch 398467 - governance - Acknowledge remove-incubated-oslo-code for Swift | 18:04 |
| timburke | clayg: it was an awful lot of typing | 18:05 |
| *** oshritf has joined #openstack-swift | 18:09 | |
| *** diogogmt has joined #openstack-swift | 18:13 | |
| *** catinthe_ has joined #openstack-swift | 18:13 | |
| *** tqtran has joined #openstack-swift | 18:13 | |
| *** catintheroof has quit IRC | 18:15 | |
| *** catintheroof has joined #openstack-swift | 18:15 | |
| *** geaaru has quit IRC | 18:16 | |
| *** oshritf has quit IRC | 18:16 | |
| *** tqtran has quit IRC | 18:18 | |
| *** catinthe_ has quit IRC | 18:18 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Add test for object updater user-agent header https://review.openstack.org/398523 | 18:21 |
| *** caiobren_ has joined #openstack-swift | 18:25 | |
| *** thebloggu has joined #openstack-swift | 18:25 | |
| *** acoles is now known as acoles_ | 18:26 | |
| *** caiobrentano_ has quit IRC | 18:28 | |
| *** Administrator_ has quit IRC | 18:32 | |
| *** Administrator_ has joined #openstack-swift | 18:32 | |
| notmyname | fungi: do you have an example of a parameterized builder? I don't know how that works | 18:35 |
| fungi | notmyname: the "contact" builder at the top of http://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/jobs/macros.yaml has parameter substitution for {name}, {email} and {irc} | 18:40 |
| notmyname | fungi: ok. how are those passed in? where are they set? | 18:41 |
| fungi | notmyname: like this job-template does: http://git.openstack.org/cgit/openstack-infra/project-config/tree/jenkins/jobs/devstack-gate.yaml#n1901 | 18:41 |
| notmyname | fungi: ah, ok. thanks | 18:43 |
| *** joeljwright has quit IRC | 18:47 | |
| *** jordanP has quit IRC | 18:48 | |
| *** chlong has joined #openstack-swift | 18:53 | |
| *** clu_ has joined #openstack-swift | 19:04 | |
| *** thebloggu has quit IRC | 19:09 | |
| *** amoralej is now known as amoralej|off | 19:12 | |
| *** tqtran has joined #openstack-swift | 19:14 | |
| *** cbartz has quit IRC | 19:23 | |
| *** jamielennox|away is now known as jamielennox | 19:26 | |
| *** ChubYann has joined #openstack-swift | 19:32 | |
| *** hseipp has joined #openstack-swift | 19:35 | |
| *** tongli has quit IRC | 19:40 | |
| *** pcaruana has quit IRC | 19:51 | |
| *** dmorita has quit IRC | 20:03 | |
| *** dmorita has joined #openstack-swift | 20:04 | |
| notmyname | fungi: why is there a "gate-{name}-tox-{envlist}-{node}" in python-jobs.yaml and also a "gate-{name}-python27-{node}" template? the only difference is the first calls tox with the envlist and the second calls python27. why isn't the -python27- wrapped into the more generic envlist one? | 20:11 |
| openstackgerrit | Tim Burke proposed openstack/swift: SLO: Concurrently HEAD segments https://review.openstack.org/391682 | 20:13 |
| notmyname | fungi: follow-up and reason for my question: can I (or why can't I) replace the 'python-jobs' entry for swift fo simply adding py27 to the envlist for the existing "gate-{name}-tox-{envlist}-{node}"? | 20:14 |
| fungi | notmyname: gate-{name}-python27 came first and explicitly called tox -e py27, then later people wanted to add a bunch of jobs calling various tox targets and so gate-{name}-tox-{envlist} was born as a means of not having that same boilerplate copied dozens of times. we could move the existing tox-based templates to use that as a cleanup measure, though that will also slightly alter the job names as a | 20:14 |
| fungi | result | 20:14 |
| fungi | investigating the second question/suggestion now | 20:15 |
| notmyname | ok | 20:15 |
| notmyname | fungi: ok, there's more than just py27 in the python-jobs group. but it seems like the rest is still tox-based. hmm..maybe not all , though | 20:23 |
| notmyname | but working through this, I can see some simplification I can make | 20:25 |
| fungi | notmyname: sorry, trying to juggle a bunch of discussions in different places at once. yes looking at swift's entry in jenkins/jobs/projects.yaml you could add a gate-{name}-tox-{envlist}-{node} variant which runs the py27 env and then use that instead of or in addition to the normal gate-swift-python27 job in zuul/layout.yaml | 20:33 |
| notmyname | fungi: cool, thanks. I'll be pushing a new patch to project config soon for the xfs stuff. thanks for your help | 20:34 |
| fungi | notmyname: though the need to be able to pass a value into the TEMPDIR envvar when invoking tox means you actually need to use a copy of that job-template with a slightly different name, and have it take a parameter for that path | 20:35 |
| fungi | but anyway, we can iterate on the review | 20:35 |
| notmyname | yeah. I think I got it passed around | 20:35 |
| *** sgundur_ has joined #openstack-swift | 20:36 | |
| fungi | notmyname: also as an outcome of our cti discussion around this, i've got this amendment going on the tc meeting agenda for next week: https://review.openstack.org/397502 | 20:38 |
| patchbot | patch 397502 - governance - Acknowledge nominal prerequisites for tests | 20:38 |
| notmyname | fungi: ah, interesting. I like that. thanks :-) | 20:39 |
| *** sgundur_ has quit IRC | 20:39 | |
| fungi | you're welcome. i dislike any appearance of hypocrisy in our policies, so happy to improve | 20:39 |
| *** sgundur_ has joined #openstack-swift | 20:40 | |
| openstackgerrit | Tim Burke proposed openstack/swift: Add X-Openstack-Request-Id to Access-Control-Expose-Headers https://review.openstack.org/397980 | 20:40 |
| *** sgundur_ has quit IRC | 20:47 | |
| *** sgundur_ has joined #openstack-swift | 20:47 | |
| *** m_kazuhiro has joined #openstack-swift | 20:53 | |
| ntata | tdasilva, yes! all the tempest plugin does is to invoke the tests through it rather than .functtests | 20:54 |
| ntata | :) | 20:54 |
| mattoliverau | morning | 20:54 |
| *** acoles_ is now known as acoles | 20:56 | |
| *** rcernin has joined #openstack-swift | 20:58 | |
| *** hseipp has quit IRC | 20:58 | |
| acoles | mattoliverau: hi | 20:58 |
| mattoliverau | acoles: hey man | 20:59 |
| notmyname | meeting time in #openstack-meeting | 20:59 |
| *** joeljwright has joined #openstack-swift | 20:59 | |
| *** ChanServ sets mode: +v joeljwright | 20:59 | |
| kota_ | morning | 20:59 |
| mattoliverau | kota_: morning | 21:00 |
| kota_ | morning mattoliverau | 21:01 |
| *** tqtran is now known as tqtran-afk | 21:01 | |
| notmyname | fungi: what's the difference between the jobs listed for a project in jenkins/jobs/projects.yaml and the ones listed under zuul/layout.yaml? | 21:05 |
| fungi | notmyname: stuff under the "jenkins" subtree is configuration fed into jenkins-job-builder, while stuff under the "zuul" subtree is configuration fed into zuul. different software with different semantics and data model | 21:09 |
| notmyname | fungi: but it seems that they basically need to mirror each other, right? | 21:09 |
| fungi | so while there is some relationship between them (mostly arbitrarily determined for consistency sake), it's an apples/oranges comparison | 21:10 |
| fungi | notmyname: well, the jenkins job configuration defines job payloads, while the zuul configuration tells the scheduler when to run them under what conditions | 21:10 |
| *** mvk has joined #openstack-swift | 21:11 | |
| fungi | so to the extent that jobs need to be defined for zuul to be able to run them (we have a ci job to catch situations where that isn't the case) there is an interrelationship yes | 21:11 |
| fungi | though constructs in jenkins configuration such as job-groups and project-templates for zuul may be similarly named even though they're not entirely analogous structures | 21:13 |
| fungi | (e.g., "jenkins-jobs" is the name of a job-group in the jenkins-job-builder configuration and the name of a project-template in the zuul layout) | 21:13 |
| fungi | er, i mean "python-jobs" | 21:13 |
| fungi | they're complimentary, but the similar names for them are merely out of convenience | 21:14 |
| fungi | it's also worth mentioning that this gets far less confusing in zuul v3, as the job playbooks and zuul layout are merged so this confusing distinction will be gone soon | 21:17 |
| fungi | it's mainly an artifact of our old use of jenkins as a job runner | 21:18 |
| *** m_kazuhiro has quit IRC | 21:23 | |
| *** caiobren_ has quit IRC | 21:24 | |
| *** caiobrentano_ has joined #openstack-swift | 21:25 | |
| *** caiobrentano_ has quit IRC | 21:29 | |
| *** vint_bra has quit IRC | 21:30 | |
| *** catinthe_ has joined #openstack-swift | 21:32 | |
| *** Jeffrey4l has quit IRC | 21:33 | |
| *** cdelatte has quit IRC | 21:34 | |
| *** m_kazuhiro has joined #openstack-swift | 21:35 | |
| *** catintheroof has quit IRC | 21:36 | |
| *** jordanP has joined #openstack-swift | 21:39 | |
| *** sams-gleb has quit IRC | 21:41 | |
| *** sams-gleb has joined #openstack-swift | 21:42 | |
| *** sams-gleb has quit IRC | 21:42 | |
| *** sams-gleb has joined #openstack-swift | 21:43 | |
| *** sams-gleb has quit IRC | 21:45 | |
| *** jordanP has quit IRC | 21:46 | |
| openstackgerrit | Merged openstack/swift: Remove duplicated 'User-Agent' header from object-updater's requests https://review.openstack.org/388999 | 21:50 |
| *** sgundur_ has quit IRC | 21:50 | |
| *** catintheroof has joined #openstack-swift | 21:51 | |
| *** catinthe_ has quit IRC | 21:51 | |
| openstackgerrit | Timur Alperovich proposed openstack/swift: Add the container_crawler library. https://review.openstack.org/398600 | 21:59 |
| tdasilva | timur: ^^^ cool! | 21:59 |
| *** sams-gleb has joined #openstack-swift | 22:01 | |
| clayg | so this warning... are we *also* going to check if you're running isa-l with parity > 4 | 22:02 |
| clayg | kota_: tdasilva: notmyname: ^ | 22:02 |
| kota_ | clayg: good perspective | 22:02 |
| jrichli | timur: nice | 22:02 |
| kota_ | clayg: probably we can validate both the libec version and # of parities at the init | 22:03 |
| clayg | and also... if you *are* using isa_l_rs_vand with parity > 4 - yeah liberasurecode 1.3.1 doesn't have a data durability issue - but the availabilty risk is still there | 22:03 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 22:03 |
| *** clu_ has quit IRC | 22:04 | |
| kota_ | clayg: but still we hold it as warn??? i think if we assume it as like a kind of Validation Error, the user already deploys the parameter cannot use it anymore. | 22:04 |
| *** clu_ has joined #openstack-swift | 22:04 | |
| clayg | kota_: that's my understanding yes, so the "warning" we eventually add to swift will not really matter the liberasure version | 22:05 |
| kota_ | clayg: cannot get the stored data, i mean. | 22:05 |
| notmyname | warning added to swift? I wouldn't think we'd do any changes there. we could, I guess, on storage policy validation | 22:05 |
| *** sams-gleb has quit IRC | 22:05 | |
| *** m_kazuhiro has quit IRC | 22:05 | |
| *** sams-gleb has joined #openstack-swift | 22:05 | |
| *** joeljwright has quit IRC | 22:05 | |
| clayg | notmyname: oh, the warning was going to be pyeclib? | 22:06 |
| notmyname | that's what I understood | 22:06 |
| kota_ | clayg, notmyname: I've thinking like as notmyname | 22:06 |
| kota_ | 've been | 22:07 |
| timur | I added mattoliverau to it too, but feel free to comment! | 22:07 |
| notmyname | (by which kota_ means that i copied him. it's his good idea) ;-) | 22:07 |
| timur | would it be helpful to also add the follow-on commit that uses this library to implements the elasticsearch integration? | 22:07 |
| clayg | ok, well same difference - my point is that we have a combination of ec_type + k + m that is never good | 22:07 |
| clayg | we should try to disallow it really | 22:08 |
| timburke | i think we definitely want a warning in swift on known-problematic storage policy configurations. it'll be sucky noise for policies already in use, but invaluable when you're brining up your POC cluster | 22:08 |
| timburke | maybe add some yes-i-know-i-screwed-up config option to silence it | 22:08 |
| clayg | timburke: i wouldn't mind requring the policy be marked deprecated | 22:08 |
| clayg | timburke: you can expose a new policy and rename this one legacy some such something w/e | 22:09 |
| tdasilva | clayg: i'm not sure i understand the problem well, so basically users should never deploy EC with m > 4 with isa-l ??? | 22:09 |
| tdasilva | and there will be no fix to whatever problem exists today? | 22:10 |
| clayg | isa_l_rs_vand, yes - kota_ is working on isa_l_rs_vand_cunch and libec 1.3.2 will allow that ec_type to work with m > 4 | 22:10 |
| clayg | tdasilva: we can fix the durability issue but not the availability issue | 22:10 |
| *** lifeless has quit IRC | 22:11 | |
| clayg | we can sorta paper over it with some crazy shit in swift that make your k+m feel like a (k+1)+(m-1) *sometimes* but it's a pretty tenuous affair | 22:11 |
| clayg | i would prefer to just deprecate support for such configurations | 22:11 |
| kota_ | tdasilva, clayg: i sent the link for the description to tdasilva in private | 22:11 |
| notmyname | clayg: +1 | 22:11 |
| tdasilva | got it | 22:11 |
| clayg | esp if we have an alternative that works with libec 1.3.2 - which seems likely | 22:12 |
| notmyname | kota_: thanks. I was just about to ;-) | 22:12 |
| *** lifeless has joined #openstack-swift | 22:12 | |
| tdasilva | clayg: +1 to your idea of deprecation support | 22:12 |
| clayg | sorry - is this whole mess still under some advisory? I'm sorta past all this | 22:12 |
| notmyname | yeah, but I just left a comment about it. https://bugs.launchpad.net/swift/+bug/1639691 | 22:13 |
| openstack | notmyname: Error: malone bug 1639691 not found | 22:13 |
| clayg | heh | 22:13 |
| notmyname | malone? | 22:13 |
| acoles | +1 for deprecation. feels like that should happen as close to source of problem as possible - e.g. the lib shouldn't allow a config that is flawed (unless some flag is passed in to force it for backwards comaptibility) | 22:14 |
| notmyname | all the patches have been open. I just want to have a sane story to tell deployers before we give them a single link to say "oh no! it's broken" | 22:14 |
| zaitcev | It says "The information on this page is private." but I can see it. How private is it? | 22:15 |
| notmyname | swift core can see it | 22:15 |
| *** sams-gleb has quit IRC | 22:15 | |
| *** sams-gleb has joined #openstack-swift | 22:15 | |
| acoles | good night | 22:16 |
| *** acoles is now known as acoles_ | 22:16 | |
| kota_ | good night acoles | 22:16 |
| kota_ | k, will try to the warning staff today, leave for break fast | 22:18 |
| clayg | notmyname: ok, so we have that - you need to upgrade to liberasurecode 1.3.1 - you need to reput objects in effected policy to repair coruption - you need to deprecate the policy and migrate data to a different policy | 22:18 |
| clayg | the different policy can either be a non isa-l policy; or isa_l_rs_vand m <= 4; or we need to review kota_'s isa_l_rs_vand_cunch and release libec 1.3.2 | 22:19 |
| clayg | it's not a great story but thems the knocks | 22:20 |
| *** sams-gleb has quit IRC | 22:20 | |
| notmyname | right. honestly, I just want that written down in a linkable place and, ideally, also point to a re-putter script | 22:20 |
| timburke | kinda makes you wish https://review.openstack.org/#/c/209329/ was landed... | 22:20 |
| patchbot | patch 209329 - swift - WIP: Changing Policies | 22:20 |
| notmyname | FYI proposed gate changes to use XFS https://review.openstack.org/#/c/394600/ | 22:21 |
| patchbot | patch 394600 - openstack-infra/project-config - enable an xfs tmpdir for swift tests | 22:21 |
| clayg | timburke: makes you wish it wasn't WIP - that's different from wishing it "was landed" ;) | 22:21 |
| timburke | ok ok, first one then the other :P | 22:22 |
| clayg | timburke: right, sorry | 22:22 |
| clayg | notmyname: well where do you want me to write it down exactly? shouldn't we just work on isa_l_rs_cauchy? | 22:23 |
| clayg | https://review.openstack.org/#/c/393263/ | 22:23 |
| patchbot | patch 393263 - liberasurecode - ISA-L Cauchy support | 22:24 |
| clayg | for me it's the next thing on my list after I merge liberasurecode packages and finish testing the reput script (which is basically the same as kota's script but tied into some other tooling we have for professional services) | 22:24 |
| notmyname | clayg: weren't we talking about opening some utils script we have a swiftstack? yeah, that. the reput one | 22:24 |
| *** dmorita has quit IRC | 22:28 | |
| *** dmorita has joined #openstack-swift | 22:28 | |
| *** dmorita has quit IRC | 22:28 | |
| *** dmorita has joined #openstack-swift | 22:28 | |
| openstackgerrit | Bryan Keller proposed openstack/swift: Send correct response for copy with invalid destination https://review.openstack.org/398631 | 22:30 |
| clayg | maybe - it's not really clear the baggage of that ecosystem in it's current state is a better solution to the problem at hand than the stand alone script kota attached to the bug report | 22:30 |
| *** sams-gleb has joined #openstack-swift | 22:31 | |
| bkeller` | acoles_: @wsgify-ing copy middleware involves changing a decent amount of it so i'll try that later | 22:32 |
| *** sams-gleb has quit IRC | 22:35 | |
| *** sams-gleb has joined #openstack-swift | 22:35 | |
| *** sams-gleb has quit IRC | 22:40 | |
| abalfour | n00b question, but since there's a +2 on https://review.openstack.org/#/c/395237 it's just waiting for a +1 Workflow to get pulled in, correct? | 22:52 |
| patchbot | patch 395237 - swift - add byteorder information and logic to ring files | 22:52 |
| *** catintheroof has quit IRC | 22:54 | |
| *** chlong has quit IRC | 22:59 | |
| notmyname | https://bugs.launchpad.net/swift/+bug/1639691 changed to public, with mitigation info added in the last comment. this should give us a single place to point people to related to libec+isa-l with more than 4 parity | 23:02 |
| openstack | Launchpad bug 1639691 in OpenStack Object Storage (swift) "EC: Swift can return corrupted Data and be able to go data lost at isa_l_rs_vand policy with >=5 parities" [Critical,Confirmed] | 23:02 |
| clayg | notmyname: yay! | 23:03 |
| clayg | abalfour: yup, looks like mattoliverau is already looking at it | 23:04 |
| *** tqtran-afk is now known as tqtran | 23:13 | |
| *** chlong has joined #openstack-swift | 23:16 | |
| *** dmorita has quit IRC | 23:18 | |
| *** dmorita has joined #openstack-swift | 23:19 | |
| abalfour | sweet, thanks clayg. | 23:21 |
| *** asettle has quit IRC | 23:23 | |
| *** dmorita has quit IRC | 23:24 | |
| *** diogogmt has quit IRC | 23:25 | |
| *** dfflanders has joined #openstack-swift | 23:25 | |
| *** chlong has quit IRC | 23:27 | |
| *** Jeffrey4l has joined #openstack-swift | 23:31 | |
| *** dmorita has joined #openstack-swift | 23:34 | |
| *** kei_yama has joined #openstack-swift | 23:35 | |
| *** dmorita has quit IRC | 23:37 | |
| *** zhugaoxiao has joined #openstack-swift | 23:38 | |
| *** Administrator_ has quit IRC | 23:40 | |
| *** dmorita has joined #openstack-swift | 23:49 | |
| *** cdelatte has joined #openstack-swift | 23:49 | |
| *** david-lyle_ has joined #openstack-swift | 23:50 | |
| *** _JZ_ has quit IRC | 23:51 | |
| *** david-lyle has quit IRC | 23:51 | |
| *** rcernin has quit IRC | 23:52 | |
| *** klamath has quit IRC | 23:53 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!