| zaitcev | I don't know about you guys, but suddenly tox started throwing errors in s3api unit tests | 00:06 |
|---|---|---|
| zaitcev | FAIL: test_object_multipart_upload_complete_failure_with_heartbeat (test.unit.co | 00:06 |
| zaitcev | mmon.middleware.s3api.test_multi_upload.TestS3ApiMultiUpload) | 00:06 |
| zaitcev | AssertionError: '<Error><Code>InvalidRequest</Code><Message>some/object: 403 Forbidden</Message><Status>400 Bad Request</Status></Error>' is not false | 00:06 |
| zaitcev | How did we stuff that through the gate, I wonder. | 00:06 |
| timburke | some trouble with the time mocking in the test? i've not seen that failure, but it sounds kind of familiar... | 00:10 |
| timburke | we *do* expect there to be a 403 Forbidden error... we just want some whitespace *first* :-/ | 00:12 |
| *** gyee has quit IRC | 01:33 | |
| *** mikecmpbll has quit IRC | 01:39 | |
| *** psachin has joined #openstack-swift | 03:01 | |
| *** openstackgerrit has joined #openstack-swift | 03:02 | |
| openstackgerrit | Pete Zaitcev proposed openstack/swift master: py3: port the container sharder https://review.openstack.org/570320 | 03:02 |
| *** gkadam-bmgr has joined #openstack-swift | 03:34 | |
| *** e0ne has joined #openstack-swift | 06:12 | |
| *** e0ne has quit IRC | 07:05 | |
| *** ccamacho has joined #openstack-swift | 07:15 | |
| *** takamatsu has joined #openstack-swift | 07:46 | |
| *** tkajinam has quit IRC | 08:14 | |
| *** mikecmpbll has joined #openstack-swift | 09:06 | |
| *** e0ne has joined #openstack-swift | 09:15 | |
| *** psachin has quit IRC | 09:40 | |
| *** jistr is now known as jistr|chat | 10:31 | |
| *** e0ne has quit IRC | 10:46 | |
| *** e0ne has joined #openstack-swift | 10:52 | |
| *** jistr|chat is now known as jistr | 11:05 | |
| *** e0ne has quit IRC | 11:24 | |
| *** mrjk_ has quit IRC | 11:33 | |
| *** e0ne has joined #openstack-swift | 11:34 | |
| *** mrjk has joined #openstack-swift | 11:35 | |
| *** nguyenhai has quit IRC | 11:56 | |
| *** nguyenhai has joined #openstack-swift | 11:56 | |
| *** nguyenhai has quit IRC | 12:00 | |
| *** nguyenhai has joined #openstack-swift | 12:00 | |
| *** ybunker has joined #openstack-swift | 12:01 | |
| *** ybunker has quit IRC | 12:14 | |
| *** e0ne has quit IRC | 13:01 | |
| *** e0ne has joined #openstack-swift | 13:04 | |
| tdasilva | rledisez: https://github.com/benfred/py-spy | 13:04 |
| *** gkadam-bmgr is now known as gkadam | 13:15 | |
| *** gkadam has quit IRC | 13:20 | |
| *** e0ne has quit IRC | 13:25 | |
| *** jistr is now known as jistr|call | 13:25 | |
| *** ybunker has joined #openstack-swift | 13:28 | |
| ybunker | I inherit the following configuration for the object-expirer.conf file (http://pasted.co/b1e486ad) | 13:29 |
| *** e0ne has joined #openstack-swift | 13:29 | |
| ybunker | and its only running on one of the proxy nodes (account and containers) are running on the data nodes | 13:29 |
| *** e0ne has quit IRC | 13:30 | |
| *** jistr|call is now known as jistr | 13:31 | |
| *** e0ne has joined #openstack-swift | 13:32 | |
| ybunker | is this correct? or it should be configured on the data nodes? and on each of the nodes?, also each node has 9 data disks for objects and 3 disks for account/container | 13:33 |
| ybunker | should i configure processes option too, right? | 13:33 |
| *** e0ne has quit IRC | 13:34 | |
| *** e0ne has joined #openstack-swift | 13:35 | |
| *** hseipp has joined #openstack-swift | 13:48 | |
| DHE | https://docs.openstack.org/swift/latest/overview_expiring_objects.html "Usually, just one instance of the swift-object-expirer daemon needs to run for a cluster" | 13:52 |
| *** pcaruana has quit IRC | 13:52 | |
| DHE | you could run it anywhere, doesn't really matter as long as it has the config and ring files it needs | 13:52 |
| *** egallen has joined #openstack-swift | 13:53 | |
| *** takamatsu has quit IRC | 13:54 | |
| ybunker | DHE: oh excellent!, thanks a lot for the help, will leave it on the proxy, regarding the processes option parameter, if i would like to not burn the disks, should i use processes = 9? | 14:01 |
| *** pcaruana has joined #openstack-swift | 14:02 | |
| tdasilva | ybunker, DHE this is still probably early still and not 100% set in stone (since it hasn't merged) but one patch that is getting more attention are related to changes to the expirer which will require it to run on the object servers. just a heads up | 14:10 |
| tdasilva | https://review.openstack.org/#/c/601950/ | 14:11 |
| patchbot | patch 601950 - swift - Enable to configure object-expirer in object-serve... - 11 patch sets | 14:11 |
| tdasilva | https://review.openstack.org/#/c/517389/40 | 14:11 |
| patchbot | patch 517389 - swift - Add object-expirer new mode to execute tasks from ... - 40 patch sets | 14:11 |
| ybunker | tdasilva, thanks for the news :) | 14:12 |
| *** e0ne has quit IRC | 14:13 | |
| kota_ | tdasilva: https://en.oui.sncf/en/ | 14:25 |
| kota_ | maybe | 14:25 |
| tdasilva | kota_: thanks! | 14:25 |
| *** e0ne has joined #openstack-swift | 14:45 | |
| *** egallen has left #openstack-swift | 14:59 | |
| *** mvkr has quit IRC | 15:00 | |
| *** mvkr has joined #openstack-swift | 15:13 | |
| tdasilva | rledisez, alecuyer: not sure you have seen this before, but it's a good read: https://bugs.launchpad.net/swift/+bug/1450656 | 15:25 |
| openstack | Launchpad bug 1450656 in OpenStack Object Storage (swift) "tmp directory causing Swift slowdown" [Undecided,Fix released] - Assigned to Shri Javadekar (shrinand) | 15:25 |
| kota_ | ok, it seems that I succeeded to build the https://github.com/alecuyer/swift/tree/master-losf environment in my saio vm, by hand. | 15:30 |
| DHE | interesting... | 15:33 |
| DHE | my upcoming swift cluster is going to make EXTENSIVE use of object expiration... | 15:34 |
| kota_ | what's up DHE? | 15:35 |
| DHE | waiting for the budget for said cluster to come through... supposed to be this quarter | 15:36 |
| kota_ | alecuyer: probably we should write down protobuf-compiler>=3 to support proto3 at https://github.com/alecuyer/swift/blob/master-losf/go/swift-rpc-losf/README.md | 15:37 |
| kota_ | even tho, we might have compiled protocol buffer files. | 15:38 |
| kota_ | interesting | 15:39 |
| * kota_ is rolling back the logs | 15:39 | |
| *** ccamacho has quit IRC | 15:41 | |
| openstackgerrit | Thiago da Silva proposed openstack/swift master: Change how O_TMPFILE support is detected https://review.openstack.org/633646 | 15:50 |
| *** ccamacho has joined #openstack-swift | 15:50 | |
| kota_ | ok, swift functional tests worked w/o 58 skipped ones. | 15:57 |
| kota_ | I'm still wondering how we could manage golang code in the python repo. | 16:19 |
| kota_ | A. maintain at the same repo and make a doc how to build/install the go code, or B. separate golang code to another repo like liberasurecode. | 16:20 |
| kota_ | either of those has pros/cons tho. | 16:20 |
| kota_ | tdasilva: perhaps, sort of what pyeclib is doing could be useful? likely, https://github.com/openstack/pyeclib/blob/master/setup.py#L78-L114 | 16:33 |
| kota_ | extend the python setup to build other language modules. | 16:34 |
| *** edausq has joined #openstack-swift | 16:34 | |
| * kota_ is not sure if current Swift can follow the way tho. | 16:35 | |
| *** e0ne has quit IRC | 16:46 | |
| *** spsurya has quit IRC | 16:54 | |
| kota_ | or perhaps, creating Makefile like as ProxyFS is doing? | 17:07 |
| kota_ | would be easy to install? | 17:07 |
| *** ccamacho has quit IRC | 17:13 | |
| *** takamatsu has joined #openstack-swift | 17:14 | |
| *** gyee has joined #openstack-swift | 17:19 | |
| *** hseipp has quit IRC | 17:26 | |
| *** mikecmpbll has quit IRC | 17:37 | |
| *** ccamacho has joined #openstack-swift | 17:53 | |
| timburke | kota_: i'd vote A, assuming the amount of work to get to either is comparable | 18:15 |
| clayg | I like the Open Infrastructure initiative/direction for the foundation - I still don't know what starlingx or airship or kata are building - but it seems like they're trying to do it in a way that doesn't reinvent the parts of open project development that aren't broken 👍 | 18:15 |
| clayg | timburke: having pyeclib and libec be separate repos was wrong in hindsight IMHO - but I think not having them in swift was good. Not sure if there's anything to learn there. | 18:17 |
| timburke | clayg: yeah, i think i agree on both those points. the lesson to learn is probably, "if these components are expected to release and upgrade together in lockstep, put them in the same repo" | 18:18 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Change how O_TMPFILE support is detected https://review.openstack.org/633646 | 18:35 |
| *** mvkr has quit IRC | 18:35 | |
| *** e0ne has joined #openstack-swift | 18:39 | |
| *** e0ne has quit IRC | 18:48 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Fix socket leak on object-server death https://review.openstack.org/575254 | 19:07 |
| openstackgerrit | Cyril Roelandt proposed openstack/swift master: Python3: fix test/unit/obj/test_replicator.py https://review.openstack.org/633645 | 19:13 |
| *** pcaruana has quit IRC | 19:30 | |
| *** ybunker has quit IRC | 19:36 | |
| *** mikecmpbll has joined #openstack-swift | 20:29 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Python3: fix test/unit/obj/test_replicator.py https://review.openstack.org/633645 | 20:33 |
| openstackgerrit | Tim Burke proposed openstack/swift master: py3: proxy some logger properties in LogAdapter https://review.openstack.org/634304 | 20:33 |
| timburke | zaitcev: idk -- the py36 job made it seem like there might've just been the two left... | 20:40 |
| timburke | as much as anything, i wanted to see whether the CPython LoggerAdapter backports would fix up the py35 job to look more like the py36 one | 20:41 |
| zaitcev | Yeah, fascinating. | 20:41 |
| timburke | i wonder who i ought to be poking to get https://github.com/python/cpython/pull/7932 landed... | 20:43 |
| timburke | so far so good on p 633645 ... py35 passed, py27 is running, and py36 is queued | 20:48 |
| patchbot | https://review.openstack.org/#/c/633645/ - swift - Python3: fix test/unit/obj/test_replicator.py - 3 patch sets | 20:48 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Address some review comments https://review.openstack.org/627574 | 20:56 |
| openstackgerrit | Tim Burke proposed openstack/swift master: s3api: fix up flakey LastModified assertions https://review.openstack.org/634318 | 21:36 |
| openstackgerrit | Tim Burke proposed openstack/swift master: docs: clean up Object Versioning page https://review.openstack.org/634320 | 21:55 |
| mattoliverau | morning | 22:02 |
| zaitcev | Geh | 22:06 |
| zaitcev | floating failures | 22:06 |
| zaitcev | mattoliverau: did you see https://review.openstack.org/570320 ? It's holding up progress in container. | 22:07 |
| patchbot | patch 570320 - swift - py3: port the container sharder - 13 patch sets | 22:07 |
| zaitcev | mattoliverau: we had some item there which was not entirely clear, like is it possible to have _own_ shard range that had other shard ranges instead of bonds? | 22:08 |
| *** ccamacho has quit IRC | 22:10 | |
| timburke | i was a little surprised by this failure: http://logs.openstack.org/54/575254/6/check/swift-probetests-centos-7/e8b151e/job-output.txt.gz#_2019-01-31_21_52_10_564943 | 22:12 |
| timburke | but then i saw https://github.com/openstack/swift/blob/2.20.0/swift/proxy/controllers/container.py#L277-L278 ... should that be HTTPServiceUnavailable instead? | 22:12 |
| mattoliverau | zaitcev: I started looking last night, but ran out of steam (lack of sleep due to 2 little monsters). I'll continue this morning after my morning stand up :) | 22:13 |
| *** tkajinam has joined #openstack-swift | 22:56 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: sharding: Make replicator logging less scary https://review.openstack.org/634342 | 23:22 |
| timburke | tdasilva: fwiw, that most-recent lower-constraints failure on p 633646 shouldn't be an issue after p 634318 | 23:27 |
| patchbot | https://review.openstack.org/#/c/633646/ - swift - Change how O_TMPFILE support is detected - 5 patch sets | 23:27 |
| patchbot | https://review.openstack.org/#/c/634318/ - swift - s3api: fix up flakey LastModified assertions - 1 patch set | 23:28 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Allow "harder" symlinks https://review.openstack.org/633094 | 23:48 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!