| seongsoocho | morning | 00:21 |
|---|---|---|
| openstackgerrit | Takashi Kajinami proposed openstack/swift master: Do not call the close method for list iterator https://review.opendev.org/759602 | 00:23 |
| kota_ | morning | 00:53 |
| * kota_ is thinking the PTG will start on Tuesday morning on Asian/Pacific TZ. | 00:53 | |
| openstackgerrit | Takashi Kajinami proposed openstack/swift master: Do not require the close method for response iterator https://review.opendev.org/759602 | 02:37 |
| openstackgerrit | Takashi Kajinami proposed openstack/swift master: Do not require the close method for response iterator https://review.opendev.org/759602 | 02:42 |
| *** rcernin has quit IRC | 03:09 | |
| *** rcernin has joined #openstack-swift | 03:14 | |
| timburke | :-/ http://lists.openstack.org/pipermail/openstack-discuss/2020-October/018287.html | 04:56 |
| timburke | ^^^ best analysis of our current gate failures, i believe | 04:56 |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-swift | 05:33 | |
| kota_ | libvirt... | 06:20 |
| *** m75abrams has joined #openstack-swift | 07:10 | |
| *** rcernin has quit IRC | 07:41 | |
| *** rpittau|afk is now known as rpittau | 07:57 | |
| *** rcernin has joined #openstack-swift | 08:01 | |
| *** rcernin has quit IRC | 08:07 | |
| *** rcernin has joined #openstack-swift | 08:23 | |
| *** rcernin has quit IRC | 08:26 | |
| *** rcernin has joined #openstack-swift | 08:33 | |
| *** rcernin has quit IRC | 09:39 | |
| *** rcernin has joined #openstack-swift | 09:59 | |
| *** rcernin has quit IRC | 10:04 | |
| *** rcernin has joined #openstack-swift | 10:23 | |
| *** rcernin has quit IRC | 10:33 | |
| *** rcernin has joined #openstack-swift | 10:37 | |
| *** rcernin has quit IRC | 10:47 | |
| *** rcernin has joined #openstack-swift | 11:37 | |
| *** rcernin has quit IRC | 11:42 | |
| *** persia_ is now known as persia | 11:52 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Add unit test for memcache error limiting with single server https://review.opendev.org/759438 | 12:28 |
| openstackgerrit | Alistair Coles proposed openstack/swift master: memcache: make minimum number of servers configurable https://review.opendev.org/759493 | 12:28 |
| openstackgerrit | Alistair Coles proposed openstack/swift master: memcache: change strategy for always available servers https://review.opendev.org/759502 | 12:28 |
| *** cschwede has joined #openstack-swift | 13:10 | |
| *** ChanServ sets mode: +v cschwede | 13:10 | |
| *** rcernin has joined #openstack-swift | 13:10 | |
| clayg | are we doing it!? Is this the PTG? let's have some coffee! | 13:28 |
| *** rcernin has quit IRC | 13:35 | |
| seongsoocho | see you soon!! | 14:44 |
| *** gyee has joined #openstack-swift | 15:04 | |
| timburke | \o/ ptg week! | 15:24 |
| timburke | reminder that we've got an etherpad going at https://etherpad.opendev.org/p/swift-ptg-wallaby and our first video session will be in ~5.5hrs | 15:25 |
| *** m75abrams has quit IRC | 15:44 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: s3api: Better handle 503s in get_container_info, too https://review.opendev.org/759735 | 16:30 |
| timburke | acoles, on p 759493 you said "I'm not sure how useful, or even desirable, it is to set minimum_servers > 1" -- should i maybe rework it all so that there's a simple yes/no config for whether error-limitng should be enabled? | 16:39 |
| patchbot | https://review.opendev.org/#/c/759493/ - swift - memcache: make minimum number of servers configurable - 2 patch sets | 16:39 |
| timburke | i feel like thre's still a case to be made that if i'm dramatically changing how i configure memcached such that each proxy is talking to a single (local) endpoint, it's preferable to have *some* attempt as using memcache rather than mobbing the db layers | 16:39 |
| timburke | if the memcache errors are manifesting as timeouts, it applies some backpressure to clients which should be helpful. if (as it was in the case that prompted me to write the patch) it's a misconfiguration that causes some actions (say, incr/decr) to fail while others are fine, it's still better to use a degraded memcached than no cache at all | 16:42 |
| timburke | 🤔 or maybe i should rework the patch to not even be configurable, and just make it so if you've got exactly one memcached server configured, we skip all error-limiting stuff... | 16:45 |
| *** dsariel has joined #openstack-swift | 16:58 | |
| *** rpittau is now known as rpittau|afk | 17:09 | |
| timburke | man, this seems to get stranger and stranger: http://lists.openstack.org/pipermail/openstack-discuss/2020-October/018304.html | 17:11 |
| timburke | i just don't quite get how we end up at "tpool + Python 3.6.x + type hinting is fubar" | 17:12 |
| acoles | timburke: I'm in favour of some kind of configurability if only to avoid even considering the change of default behaviour. I think I also mention of making the error limit parameters configurable (or did I imagine that?) in which case that may be sufficient i.e. for your case the error limit duration could be set to zero | 17:13 |
| acoles | s/also mention/also saw mention/ | 17:14 |
| timburke | yeah, pretty sure mattoliverau saw that as an immediate red flag that they were all hardcoded | 17:14 |
| acoles | timburke: do you have a schedule of topics for each swift PTG session? or are we going to agree that in the first session? | 18:25 |
| timburke | no formal schedule yet -- i think it'd be a good thing to kinda work out in the first session | 18:25 |
| timburke | at the very least, pick a slot for some ops-feedback so i can get an announcement out to the ML | 18:26 |
| *** tdasilva has joined #openstack-swift | 19:55 | |
| *** ChanServ sets mode: +v tdasilva | 19:55 | |
| mattoliverau | Morning, so just to confirm start time, off schedule says 22 utc, but live schedule says 21 utc. So thought I'd confirm which is correct. | 20:00 |
| mattoliverau | *PDF shedule | 20:00 |
| timburke | should be 2100 UTC | 20:02 |
| timburke | and good morning, mattoliverau! | 20:03 |
| seongsoocho | morning ! | 20:10 |
| *** tdasilva_ has joined #openstack-swift | 20:28 | |
| *** ChanServ sets mode: +v tdasilva_ | 20:28 | |
| *** tdasilva has quit IRC | 20:31 | |
| *** samueldmq has quit IRC | 20:45 | |
| *** alecuyer has quit IRC | 20:45 | |
| *** samueldmq has joined #openstack-swift | 20:45 | |
| *** alecuyer has joined #openstack-swift | 20:45 | |
| *** rcernin has joined #openstack-swift | 21:00 | |
| zaitcev | so, what's the Zoom ID? | 21:00 |
| timburke | ptg time! https://www.openstack.org/ptg/rooms/kilo | 21:00 |
| zaitcev | oh | 21:00 |
| *** rcernin has quit IRC | 21:01 | |
| *** rcernin has joined #openstack-swift | 21:01 | |
| *** rcernin has quit IRC | 21:10 | |
| timburke | ptg etherpad: https://etherpad.opendev.org/p/swift-ptg-wallaby | 21:19 |
| timburke | clayg, the handoff table idea also makes me think of https://review.opendev.org/#/c/740866/2/swift/common/ring/ring.py | 21:25 |
| patchbot | patch 740866 - swift - ring: Be willing to query nodes with no assignments - 2 patch sets | 21:25 |
| *** rcernin has joined #openstack-swift | 21:35 | |
| *** rcernin has quit IRC | 21:43 | |
| timburke | this discussion also reminds me of https://review.opendev.org/#/c/670674/ | 21:47 |
| patchbot | patch 670674 - swift - WIP: Allow ring lookups via a service - 2 patch sets | 21:47 |
| ormandj | https://docs.openstack.org/swift/latest/overview_ring.html#history third paragraph was the reasoning "latency" | 21:49 |
| ormandj | i definitely think it's worth revisiting, times have changed :) | 21:50 |
| *** rcernin has joined #openstack-swift | 22:00 | |
| *** rcernin has quit IRC | 22:06 | |
| *** rcernin has joined #openstack-swift | 22:22 | |
| *** rcernin has quit IRC | 22:22 | |
| *** rcernin has joined #openstack-swift | 22:22 | |
| *** rcernin has quit IRC | 22:32 | |
| *** rcernin has joined #openstack-swift | 22:33 | |
| *** timur has quit IRC | 22:42 | |
| *** timur has joined #openstack-swift | 22:50 | |
| *** tkajinam has quit IRC | 22:55 | |
| *** tkajinam has joined #openstack-swift | 22:55 | |
| timburke | thanks ormandj! | 23:08 |
| timburke | ha! "ring changes are currently infrequent enough that manual control was sufficient" | 23:08 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!