| *** aselius has quit IRC | 00:11 | |
| *** silor has joined #openstack-swift | 00:14 | |
| *** wasmum has quit IRC | 00:15 | |
| *** mingyu has joined #openstack-swift | 00:18 | |
| *** mingyu has quit IRC | 00:22 | |
| *** itlinux has joined #openstack-swift | 00:24 | |
| zaitcev | timburke: please feel free to explain to me why I'm a moron, but both of them looked strange! I'll be back from a class in a few hours. | 00:29 |
|---|---|---|
| timburke | zaitcev: that's the way i did it! parent commit of https://review.openstack.org/#/c/466952/2 is https://review.openstack.org/#/c/448240/12 | 00:43 |
| patchbot | patch 466952 - swift - Follow-up for per-policy proxy configs | 00:43 |
| patchbot | patch 448240 - swift - Enable per policy proxy config options | 00:43 |
| zaitcev | Oooh | 00:44 |
| zaitcev | The "Related Changes" was beyond the screen on the right | 00:44 |
| notmyname | timburke: my gut reaction to patch 466953 is the same as zaitcev's. I'm looking forward to your comment on it | 00:44 |
| patchbot | https://review.openstack.org/#/c/466953/ - swift - Remove tempauth docs from deployment guide | 00:44 |
| zaitcev | In pennance I'll review that other one too, unless someone lands it sooner | 00:44 |
| timburke | ah, yeah... i feel like the review screen has been feeling wider, lately... wasn't sure if it was just me and my custom dashboards, though | 00:44 |
| notmyname | zaitcev: the per-policy configs? that one, I hope will be landing tomorrow | 00:45 |
| zaitcev | yes | 00:46 |
| notmyname | I mean, feel free to look at it, but I expect it to land Real Soon Now(tm) anyway. there's been a lot of back and forth on it already | 00:46 |
| notmyname | zaitcev: where do we stand on https://review.openstack.org/#/c/427911/ ? just looks liek you're waiting for reviews | 00:47 |
| patchbot | patch 427911 - swift - PUT+POST and its development test | 00:47 |
| zaitcev | got it. I'm off to the dojo now, I'll look at the patch unless I break my neck today | 00:47 |
| notmyname | zaitcev: if so, I'll bug some people about it | 00:47 |
| notmyname | zaitcev: have a good evening :-) | 00:47 |
| zaitcev | (notmyname, yes, I'm waiting for reviews. thanks... I hoped Clay/Sam/Alasdair would come down on the idea and consider if I'm actually fixing the issues) | 00:50 |
| *** bob_cheesey has joined #openstack-swift | 00:50 | |
| *** tovin07_ has joined #openstack-swift | 00:54 | |
| *** alenavolk has joined #openstack-swift | 00:57 | |
| *** alenavolk has quit IRC | 00:58 | |
| *** itlinux has quit IRC | 01:06 | |
| kota_ | good morning | 01:21 |
| *** silor has quit IRC | 01:28 | |
| *** JimCheung has quit IRC | 01:29 | |
| *** JimCheung has joined #openstack-swift | 01:30 | |
| *** JimCheung has quit IRC | 01:34 | |
| *** itlinux has joined #openstack-swift | 01:35 | |
| *** abhitechie has joined #openstack-swift | 01:46 | |
| *** abhitechie has quit IRC | 01:52 | |
| *** itlinux has quit IRC | 01:52 | |
| *** abhitechie has joined #openstack-swift | 01:53 | |
| *** abhitechie has quit IRC | 01:54 | |
| *** abhitechie has joined #openstack-swift | 01:54 | |
| *** bob_cheesey has left #openstack-swift | 01:56 | |
| *** zhurong has joined #openstack-swift | 02:11 | |
| *** mingyu has joined #openstack-swift | 02:13 | |
| *** klrmn has quit IRC | 02:16 | |
| clayg | kota_: is the per-policy config stuff going to land today!? Is it awesome!!!? (genuinely don't know; haven't looked at it; just curious) | 02:19 |
| kota_ | clayg: it's going to be good shape imo. I had some comments yesterday, acoles seemd to do great work while I was asleep. Not sure if that results in landing today. | 02:21 |
| kota_ | but timburke looks already add his +2 for that, so i hope no big blocker exists there | 02:22 |
| *** klrmn has joined #openstack-swift | 02:35 | |
| *** alenavolk has joined #openstack-swift | 02:38 | |
| clayg | wow, great work! good luck. | 02:41 |
| *** alenavolk has quit IRC | 02:42 | |
| *** klrmn has joined #openstack-swift | 02:50 | |
| *** _JZ_ has joined #openstack-swift | 02:56 | |
| *** SkyRocknRoll has joined #openstack-swift | 03:02 | |
| notmyname | rledisez: (question for an async answer) did you get any feedback on https://etherpad.openstack.org/p/swift-losf-meta-storage this week? | 03:09 |
| *** itlinux has joined #openstack-swift | 03:10 | |
| *** mingyu has quit IRC | 03:17 | |
| *** alenavolk has joined #openstack-swift | 03:20 | |
| *** abhitechie has quit IRC | 03:39 | |
| clayg | notmyname: it's weird reading some of the comments there - it's not clear to me everyone commenting is having the same conversation? | 03:41 |
| clayg | or perhaps rather it's just not obvious to me how having object metadata in the KV index effects index rebuild time | 03:42 |
| clayg | well... maybe the conversation got side tracked on rebuild because the "replicate all metadata in the volume" design can at least in theory rebuild the entire KV index from the volume - which is one of the enabling features that allows the design to flush/sync the kv index to disk periodically instead of after every update. | 03:44 |
| clayg | perhaps the implementation that kept some object metadata *only* in KV index didn't plan to support the case where the KV index was lost because they didn't consider it could be out of sync with the volume after crash? | 03:45 |
| *** alenavolk has quit IRC | 03:53 | |
| *** winggundamth has joined #openstack-swift | 03:57 | |
| *** kei_yama has quit IRC | 04:00 | |
| *** kei_yama has joined #openstack-swift | 04:01 | |
| *** SkyRocknRoll has quit IRC | 04:02 | |
| *** klrmn has quit IRC | 04:04 | |
| *** itlinux has quit IRC | 04:08 | |
| *** psachin has joined #openstack-swift | 04:09 | |
| *** ChanServ changes topic to "Swift is 7 years old! | Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/ | Meetings: https://wiki.openstack.org/wiki/Meetings/Swift | Review Dashboard: http://not.mn/reviews.html" | 04:09 | |
| -openstackstatus- NOTICE: Sufficient free space has been reclaimed that jobs are passing again; any POST_FAILURE results can now be rechecked. | 04:09 | |
| *** zhurong has quit IRC | 04:11 | |
| *** zhurong has joined #openstack-swift | 04:15 | |
| *** chsc has joined #openstack-swift | 04:15 | |
| *** mingyu has joined #openstack-swift | 04:18 | |
| *** chsc has quit IRC | 04:20 | |
| *** lucasxu has joined #openstack-swift | 04:20 | |
| *** mingyu has quit IRC | 04:22 | |
| *** psachin has quit IRC | 04:35 | |
| *** psachin has joined #openstack-swift | 04:53 | |
| *** bkopilov has joined #openstack-swift | 04:55 | |
| *** jaosorior_away is now known as jaosorior | 04:55 | |
| *** lucasxu has quit IRC | 05:10 | |
| *** winggundamth has quit IRC | 05:12 | |
| *** links has joined #openstack-swift | 05:20 | |
| zaitcev | I looked at the policy section thing and the abuse of __file__ felt so wrong. But then Tim & Kota approve, so... | 05:25 |
| zaitcev | And it can already by a directory | 05:25 |
| zaitcev | s/by /be / | 05:25 |
| clayg | pretty sure we have a few places were we safely assume config dicts returned from readconf have __file__ - not sure on the context | 05:26 |
| zaitcev | It's still there, but it's not a string anymore. | 05:27 |
| clayg | *whoa* | 05:31 |
| zaitcev | I promised to look at https://review.openstack.org/#/c/448240 today, but I just can't say anything smart about it. | 05:32 |
| patchbot | patch 448240 - swift - Enable per policy proxy config options | 05:32 |
| zaitcev | timburke, about teaching users not to use TempAuth | 05:33 |
| zaitcev | timburke, http://zaitcev.livejournal.com/150962.html | 05:34 |
| zaitcev | timburke, I'm not Russian enough anymore in that context. | 05:34 |
| *** chsc has joined #openstack-swift | 05:43 | |
| *** chsc has quit IRC | 05:48 | |
| *** _JZ_ has quit IRC | 05:52 | |
| *** psachin has quit IRC | 05:53 | |
| *** rcernin has quit IRC | 06:00 | |
| *** psachin has joined #openstack-swift | 06:05 | |
| *** cshastri has joined #openstack-swift | 06:08 | |
| *** mingyu has joined #openstack-swift | 06:11 | |
| clayg | Them: tell the customer he is stupid; here, take N 😍 | 06:13 |
| *** gyee has quit IRC | 06:17 | |
| *** rcernin has joined #openstack-swift | 06:17 | |
| *** psachin has quit IRC | 06:23 | |
| *** psachin has joined #openstack-swift | 06:23 | |
| *** jeffli has joined #openstack-swift | 06:25 | |
| *** hoonetorg has quit IRC | 06:25 | |
| *** ChubYann has quit IRC | 06:27 | |
| *** hoonetorg has joined #openstack-swift | 06:37 | |
| *** mingyu has quit IRC | 06:50 | |
| *** mingyu has joined #openstack-swift | 06:51 | |
| *** mingyu has quit IRC | 06:51 | |
| *** bkopilov has quit IRC | 06:55 | |
| *** mingyu has joined #openstack-swift | 06:55 | |
| *** mingyu has quit IRC | 06:59 | |
| *** bkopilov has joined #openstack-swift | 07:08 | |
| *** jroll has quit IRC | 07:12 | |
| *** pcaruana has joined #openstack-swift | 07:16 | |
| *** mingyu has joined #openstack-swift | 07:23 | |
| kota_ | zaitcev: I suppose the __file__ thing is still under discussion, https://review.openstack.org/#/c/448240/11/swift/common/wsgi.py@118 | 07:29 |
| patchbot | patch 448240 - swift - Enable per policy proxy config options | 07:29 |
| *** skudlik has joined #openstack-swift | 07:33 | |
| *** hseipp has joined #openstack-swift | 07:34 | |
| *** mingyu has quit IRC | 07:46 | |
| rledisez | notmyname: about losf, we got feedback on meta storage on the etherpad. we also discussed a lot with alecuyer and jeffli and we settled on a volume format. meta will be stored in volume | 07:47 |
| rledisez | notmyname: we now need to work on K/V format (rocksdb vs leveldb), format of key and value | 07:47 |
| rledisez | notmyname: alecuyer did some preliminary bench on 1 big K/V vs multiple small K/V. it seems the memory overhead of having multiple small K/V would be a major problem | 07:48 |
| acoles | good morning | 07:48 |
| acoles | kota_: thanks for recheck | 07:49 |
| clayg | I think the __file__ StringIO is fine - it's not ideal - particularly the paste handling - but that was pre-existing. | 07:50 |
| clayg | the readconf stuff is totally find - entirely explicit - you could add something to couple ConfigString with readconf it it made it easier to reason about ... or you can just leaverge the fact we're already looking at __file__ and readconf already does the right thing with readables - done | 07:51 |
| clayg | now that PasteDeploy is unmaintained we should inline/rewrite the parts we use and it'll probably be a lot easier to follow! | 07:52 |
| acoles | clayg: I don't love that solution myself, but IMHO the ConfigString thing was already a kludge, just setting __file__ to arbitrary and fairly useless value 'string' | 07:54 |
| clayg | yeah that's my opinion as well | 07:55 |
| kota_ | acoles: good morning | 07:55 |
| acoles | clayg: and IIRC we already rely on __file__ being there, as you said, elsewhere | 07:57 |
| clayg | I think people just see the ConfigString stuff and go like "wtf how is this plumbed through to paste's loadconext business" - but your change doesn't really have anything to do with that mess/garbage - all that paste stuff is impossible to hold in your head | 07:57 |
| clayg | yeah __file__ is fully legit - how else you gunna re-read the parts of the config you missed? | 07:58 |
| clayg | is that really the most gross only wtf thing at this point!? That's *awesome*! | 08:00 |
| *** cbartz has joined #openstack-swift | 08:00 | |
| *** hseipp has quit IRC | 08:02 | |
| acoles | clayg: in copy middleware for example, and some of the daemons that go looking for other conf sections, we use __file__ there | 08:02 |
| acoles | clayg: I think timburke has some great follow up to make pieces a little cleaner. somehow I missed his patch yesterday so independently think I got most but not all of his changes in there | 08:04 |
| clayg | yeah using __file__ is ubiquitous | 08:04 |
| *** oshritf has joined #openstack-swift | 08:04 | |
| kota_ | general thinking of the context "re-open", the __file__ should be something re-opanable but currently swift allows the __file__ as an file-like object. | 08:15 |
| kota_ | the magic is in swift.common.utils.readconf | 08:15 |
| acoles | kota_: in paste-deploy loadwsgi the __file__ is abspath to a file, but swift has already abused that in ConfigString by setting __file__='string' (which is not a path to a file!) - my logic is that making file a StringIO is no worse than it being an non-existent arbitrary 'path' that cannot be opened | 08:25 |
| *** joeljwright has joined #openstack-swift | 08:25 | |
| *** ChanServ sets mode: +v joeljwright | 08:25 | |
| acoles | especially since readconf can read from a StringIO, so there is at least some cosnsitent 'design' | 08:26 |
| acoles | consistent* | 08:26 |
| kota_ | acoles: exactly | 08:27 |
| kota_ | acoles: i'm just looking at why we need ConfigString, such a abused one :/ | 08:28 |
| acoles | kota_: AFAICT it was added to support the hardcoded default InternalClient config in container sync, https://github.com/openstack/swift/blob/671254224a4a4710e7556535ee68bd999536ab8d/swift/container/sync.py#L230-L230 | 08:29 |
| clayg | ConfigString is there for backwards compat when some something container sync something internal client something default interal_client.conf | 08:29 |
| clayg | or... that | 08:29 |
| kota_ | i don't want to mess InternalClient/container sync with just an idea but it would be needed to refactor/correct the impl... I'm feeling | 08:29 |
| acoles | ...yeah, that | 08:29 |
| kota_ | oh, that for backward compat... | 08:30 |
| acoles | The hard coded conf is there to save ops needing to create an internal client conf file when the change from direct to internal client was made in container sync i.e. it is an upgrade thing | 08:30 |
| acoles | kota_: yeah, backwards compat, don't break on upgrade | 08:31 |
| clayg | ah... yeah it was a superb hack I'm afraid - the PasteDeploy docs were never that great when you could find them - I had to troll through the paste source just to get *something* to work | 08:31 |
| kota_ | but I'm wondering why we could make it as default values in the InternalClient... or preparing *conf dict* for InternalClient init | 08:31 |
| kota_ | though InternalClient take a path for init though | 08:32 |
| clayg | pretty sure you can't "fix" it w/o inline some more/all of PasteDeploy - which would be awesome! | 08:32 |
| kota_ | right now. | 08:32 |
| clayg | kota_: pipeline loading :'( | 08:32 |
| kota_ | clayg: that's too bad :( | 08:32 |
| clayg | you have to go through paste if you want to get a pipeline - no readconf can make that happen | 08:32 |
| kota_ | i see | 08:33 |
| acoles | kota_: would be nice but it's not just a single conf dict, it's an entire pipeline plus filter etc | 08:33 |
| clayg | I mean - we *could* - it's just entrypoints! it would be *amazing* | 08:33 |
| kota_ | hmm.... | 08:34 |
| clayg | you could have one code path that reads internal_client.conf with paste - and a totally different path that builds the backwards compat client from primitives by hand | 08:36 |
| clayg | but at *this* point there's stuff like this floating around -> https://gist.github.com/clayg/5ab6001c13a733ae23b0fdf905af2a60 | 08:36 |
| * clayg shrugs | 08:37 | |
| clayg | it's old cruft - the ugly parts of ConfigString aren't acoles fault | 08:37 |
| acoles | actually, since the default internal client conf is *fixed*, we could maybe just instantiate the filters/app directly and not even need a conf parsing step? something like ic = CatchErrors(ProxyLogging(MemCache(Application))) | 08:38 |
| acoles | I'm sure if i typed the code I'd realise the flaw in my thinking | 08:38 |
| acoles | but whatever, that would be a different patch | 08:38 |
| kota_ | acoles: self.app =, exactly it should be a different patch | 08:39 |
| clayg | acoles: i find it *fascinating* that the __file__ key was there at all? I thought removing it on master would cause a unittest to fail or something... | 08:39 |
| clayg | You can code it up and see if you like it - but I'm guessing you | 08:39 |
| clayg | will find that it's easier to maintain the inlined config file | 08:40 |
| kota_ | clayg: at least, test_wsgi.py didn't fail, I tried to run unit | 08:40 |
| clayg | similar reason we stopped having inprocess functests build pipelines by hand and they instead just load the example configs | 08:40 |
| kota_ | when removing __file__ in ConfigString | 08:40 |
| clayg | ... i could be wrong | 08:40 |
| acoles | clayg: I won't be doing that *today* ;) ... like, replace something that works with something else that works :P | 08:40 |
| kota_ | ah, inprocess functests | 08:41 |
| clayg | i just use them as an example since they used to do the app = Filter(Filter(Filter(app))) trick | 08:42 |
| acoles | clayg: you mean that __file__ key is set in ConfigString - yeah it does seem like maybe it wasn't even necessary | 08:44 |
| acoles | test_sync.py all pass if i remove the __file__ | 08:46 |
| kota_ | but with the config per policy needs the attribute yeah | 08:46 |
| kota_ | s/with// | 08:46 |
| kota_ | acoles: lol | 08:47 |
| acoles | kota_: yes and in the patch the new proxy/test_server.py:TestProxyServerConfigStringLoading tests will fail if ConfigString __file__ is not set | 08:51 |
| acoles | but on master seems like nothing cares about the __file__ value, which makes sense because it is arbitrary | 08:51 |
| *** kei_yama has quit IRC | 08:52 | |
| *** mvk has joined #openstack-swift | 08:58 | |
| *** hseipp has joined #openstack-swift | 09:20 | |
| *** mvk has quit IRC | 09:42 | |
| *** links has quit IRC | 09:55 | |
| *** tristanC has joined #openstack-swift | 09:56 | |
| *** adriant has quit IRC | 09:58 | |
| *** skudlik has quit IRC | 10:02 | |
| *** tovin07_ has quit IRC | 10:03 | |
| *** zhurong has quit IRC | 10:04 | |
| *** links has joined #openstack-swift | 10:08 | |
| *** mvk has joined #openstack-swift | 10:14 | |
| *** jroll has joined #openstack-swift | 10:25 | |
| *** skudlik has joined #openstack-swift | 11:10 | |
| *** MVenesio has joined #openstack-swift | 11:45 | |
| *** early has quit IRC | 11:49 | |
| *** early has joined #openstack-swift | 11:49 | |
| *** MVenesio has quit IRC | 11:59 | |
| *** jeffli has quit IRC | 12:09 | |
| *** lucasxu has joined #openstack-swift | 12:31 | |
| *** klamath has joined #openstack-swift | 12:32 | |
| *** klamath has joined #openstack-swift | 12:32 | |
| *** NM has joined #openstack-swift | 12:33 | |
| *** zhurong has joined #openstack-swift | 12:35 | |
| kota_ | hmm.... it's close to time over to stay my office... | 12:35 |
| *** zhurong has quit IRC | 12:36 | |
| acoles | kota_: yeah it must be late there | 12:36 |
| kota_ | acoles: sorry, i didn't finish the work to review yet. I'll be back to the patch tomorrow but if another core makes sure the patch, it's ok to land it. currently I didn't hit significant issue on that patch. | 12:41 |
| acoles | kota_: ok, thanks for working on it | 12:41 |
| *** stradling has quit IRC | 12:43 | |
| *** stradling has joined #openstack-swift | 12:43 | |
| *** lucasxu has quit IRC | 12:50 | |
| *** MVenesio has joined #openstack-swift | 13:09 | |
| *** lucasxu has joined #openstack-swift | 13:20 | |
| *** NM has quit IRC | 13:29 | |
| *** NM has joined #openstack-swift | 13:35 | |
| *** jeffli has joined #openstack-swift | 13:55 | |
| *** alenavolk has joined #openstack-swift | 13:55 | |
| *** jeffli has quit IRC | 14:00 | |
| *** mingyu has joined #openstack-swift | 14:08 | |
| *** alenavolk has quit IRC | 14:12 | |
| *** mingyu has quit IRC | 14:20 | |
| *** NM has quit IRC | 14:21 | |
| *** vint_bra has joined #openstack-swift | 14:23 | |
| *** NM has joined #openstack-swift | 14:25 | |
| *** vint_bra has quit IRC | 14:35 | |
| *** lucasxu has quit IRC | 14:36 | |
| *** lucasxu has joined #openstack-swift | 14:37 | |
| *** chlong has joined #openstack-swift | 14:38 | |
| *** itlinux has joined #openstack-swift | 14:45 | |
| *** vint_bra has joined #openstack-swift | 14:49 | |
| *** lucasxu has quit IRC | 14:50 | |
| *** vint_bra has quit IRC | 14:50 | |
| *** cshastri has quit IRC | 14:50 | |
| *** lucasxu has joined #openstack-swift | 14:52 | |
| *** aselius has joined #openstack-swift | 14:53 | |
| *** psachin has quit IRC | 14:55 | |
| *** bob_cheesey has joined #openstack-swift | 15:00 | |
| *** bob_cheesey has quit IRC | 15:06 | |
| *** vint_bra has joined #openstack-swift | 15:07 | |
| *** bob_cheesey has joined #openstack-swift | 15:11 | |
| timburke | good morning | 15:12 |
| *** bob_cheesey has joined #openstack-swift | 15:14 | |
| *** gyee has joined #openstack-swift | 15:15 | |
| *** bob_cheesey has quit IRC | 15:15 | |
| *** rcernin has quit IRC | 15:18 | |
| *** itlinux has quit IRC | 15:33 | |
| *** klrmn has joined #openstack-swift | 15:35 | |
| *** itlinux has joined #openstack-swift | 15:36 | |
| *** MVenesio has quit IRC | 15:36 | |
| notmyname | good morning | 15:45 |
| *** mvk_ has joined #openstack-swift | 15:50 | |
| *** Sukhdev_ has joined #openstack-swift | 15:50 | |
| *** skudlik has quit IRC | 15:51 | |
| *** itlinux has quit IRC | 15:52 | |
| *** stew925 has joined #openstack-swift | 15:55 | |
| *** lucasxu has quit IRC | 15:56 | |
| notmyname | bouncing my bouncer... | 15:57 |
| *** notmyname has quit IRC | 15:57 | |
| *** patchbot has quit IRC | 15:57 | |
| *** patchbot has joined #openstack-swift | 15:59 | |
| *** notmyname has joined #openstack-swift | 16:01 | |
| *** ChanServ sets mode: +v notmyname | 16:01 | |
| notmyname | hello again | 16:02 |
| *** cbartz has quit IRC | 16:09 | |
| notmyname | note the proposed schedule for the Denver PTG | 16:10 |
| notmyname | https://docs.google.com/spreadsheets/d/1xmOdT6uZ5XqViActr5sBOaz_mEgjKSCY7NEWcAEcT-A/pubhtml?gid=397241312&single=true | 16:10 |
| notmyname | also, if you're running a swift cluster, your feedback is invaluable. based on the experiences at the first PTG and Forum, I'd strongly recommend that ops attend the PTG instead of the summit/forum | 16:11 |
| *** klrmn has quit IRC | 16:15 | |
| *** oshritf_ has joined #openstack-swift | 16:17 | |
| notmyname | any ansible folks around? | 16:18 |
| *** oshritf has quit IRC | 16:20 | |
| *** chsc has joined #openstack-swift | 16:23 | |
| *** chsc has joined #openstack-swift | 16:23 | |
| *** pcaruana has quit IRC | 16:24 | |
| *** bkopilov has quit IRC | 16:25 | |
| *** JimCheung has joined #openstack-swift | 16:26 | |
| *** alenavolk has joined #openstack-swift | 16:27 | |
| *** chsc has quit IRC | 16:32 | |
| *** alenavolk has quit IRC | 16:33 | |
| *** bkopilov has joined #openstack-swift | 16:34 | |
| *** esnyder has joined #openstack-swift | 16:35 | |
| *** _JZ_ has joined #openstack-swift | 16:38 | |
| *** hseipp has quit IRC | 16:45 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Follow-up for per-policy proxy configs https://review.openstack.org/466952 | 16:50 |
| *** lucasxu has joined #openstack-swift | 16:50 | |
| *** mingyu has joined #openstack-swift | 16:50 | |
| *** Sukhdev_ has quit IRC | 16:57 | |
| *** mingyu has quit IRC | 16:58 | |
| *** alenavolk has joined #openstack-swift | 17:01 | |
| *** lucasxu has quit IRC | 17:03 | |
| *** lucasxu has joined #openstack-swift | 17:07 | |
| *** _JZ_ has quit IRC | 17:08 | |
| *** chsc has joined #openstack-swift | 17:10 | |
| *** chsc has quit IRC | 17:11 | |
| *** chsc has joined #openstack-swift | 17:11 | |
| *** klrmn has joined #openstack-swift | 17:13 | |
| *** links has quit IRC | 17:16 | |
| *** _JZ_ has joined #openstack-swift | 17:17 | |
| *** _JZ_ has quit IRC | 17:17 | |
| *** mvk_ has quit IRC | 17:17 | |
| *** _JZ_ has joined #openstack-swift | 17:19 | |
| *** ChubYann has joined #openstack-swift | 17:21 | |
| *** mvk has quit IRC | 17:22 | |
| *** jaosorior has quit IRC | 17:24 | |
| *** mingyu has joined #openstack-swift | 17:24 | |
| tdasilva | notmyname: what's your ansible question? | 17:28 |
| *** oshritf_ has quit IRC | 17:29 | |
| notmyname | "why doesn't it work?" ;-) | 17:30 |
| tdasilva | lol | 17:31 |
| notmyname | headed to a meeting now, but I'm having struggles with the mount module. everything seems to work during the ansible run, but the drives aren't mounted when I log into the guest | 17:31 |
| notmyname | anyway, i'll be back on it after this meeting | 17:32 |
| tdasilva | k, will link something i have done before that might help | 17:32 |
| notmyname | thanks | 17:32 |
| *** logan- has quit IRC | 17:34 | |
| tdasilva | notmyname: not sure this helps, but maybe it's worth as reference material: | 17:35 |
| tdasilva | https://github.com/thiagodasilva/ansible-saio/blob/master/roles/swift/tasks/main.yml#L16 | 17:35 |
| tdasilva | https://github.com/thiagodasilva/swift-ansible/blob/master/roles/prepare_disks/tasks/main.yml | 17:35 |
| *** Sukhdev_ has joined #openstack-swift | 17:46 | |
| *** mvk_ has joined #openstack-swift | 17:53 | |
| mwheckmann | notmyname: going into a meeting too, but you can shoot some Ansible questions my way too. | 17:58 |
| notmyname | good news! I figured out it's not ansible (directly) that's causing problems | 18:01 |
| notmyname | thanks mwheckmann and tdasilva | 18:02 |
| notmyname | hmm... so now I have an lxd question | 18:08 |
| notmyname | I can `lxc start <container>`, but that doesn't remount what's listed in /etc/fstab. why not? is there some normal boot process that it is *not* doing? is there a different way to start it? | 18:08 |
| *** alenavolk has quit IRC | 18:17 | |
| *** oshritf_ has joined #openstack-swift | 18:21 | |
| *** oshritf_ has quit IRC | 18:30 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Clean up some assertions in test_reconstructor https://review.openstack.org/467710 | 18:33 |
| timburke | acoles (and mattoliverau, when you come online): now that we've got green balls on https://review.openstack.org/#/c/450491/, any opposition to me squashing it into the primary sharding patch? | 18:36 |
| patchbot | patch 450491 - swift - fixup! Add container sharding to Swift containers | 18:36 |
| *** alenavolk has joined #openstack-swift | 18:37 | |
| *** itlinux has joined #openstack-swift | 18:52 | |
| timburke | clayg: thanks, i was just looking at that failure, too. i guess i went a bit overboard on how much i wanted to assert when we weren't being terrible specific before | 18:57 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Clean up some assertions in test_reconstructor https://review.openstack.org/467710 | 18:57 |
| clayg | oh was that it? I figured it was just pre-existing order coupling that gave way when the expected sets were built slightly different. | 18:58 |
| timburke | i also scaled back how much of the node dict to verify, since that seemed preferable to bumping maxDiff (i think?) | 18:59 |
| clayg | idk? maxDiff only matters when it fails ;) | 18:59 |
| timburke | yeah, it was stuff like https://review.openstack.org/#/c/467710/1/test/unit/obj/test_reconstructor.py@1148 where we were previously only saying `self.assertEqual(len(found_jobs), 6)` | 19:01 |
| patchbot | patch 467710 - swift - Clean up some assertions in test_reconstructor | 19:01 |
| *** itlinux has quit IRC | 19:01 | |
| *** alenavolk has quit IRC | 19:04 | |
| timburke | ah, right, and then the remaining failures are because we shuffle the suffixes to hash, i guess? | 19:08 |
| clayg | i don't know timburke ... i don't know | 19:13 |
| timburke | nope, found it: also dict ordering | 19:13 |
| clayg | checking out patch 448240 | 19:13 |
| patchbot | https://review.openstack.org/#/c/448240/ - swift - Enable per policy proxy config options | 19:13 |
| clayg | @timburke was patch set 2 still failing for you? (!) | 19:14 |
| *** itlinux has joined #openstack-swift | 19:15 | |
| timburke | clayg: i ran it a couple times, but since it's a 50/50 chance each run... not enough to pop like the gate did | 19:15 |
| clayg | patchset 1 was really consistently passing for me w/o the `PYTHONHASHSEED=random` | 19:16 |
| clayg | and failing reliably with it - patchset 2 was solid OMM | 19:16 |
| clayg | maybe we could plumb that into our unit-test run somehow? Or maybe the gate/tox already does it!? | 19:17 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Clean up some assertions in test_reconstructor https://review.openstack.org/467710 | 19:22 |
| *** itlinux has quit IRC | 19:23 | |
| timburke | clayg: i think it must already? or something: http://logs.openstack.org/10/467710/2/check/gate-swift-tox-xfs-tmp-py27-ubuntu-xenial/f713c45/console.html#_2017-05-24_19_00_48_568355 | 19:23 |
| timburke | yeah.... https://github.com/tox-dev/tox/blob/2.7.0/CHANGELOG#L471-L474 had it for years | 19:26 |
| timburke | i'm loving that i got around to doing https://review.openstack.org/#/c/455439/ though -- makes it cheap to throw up, review, and merge this sort of a patch | 19:29 |
| patchbot | patch 455439 - openstack-infra/project-config - swift: Skip long-running dsvm jobs for unrelated c... (MERGED) | 19:29 |
| *** itlinux has joined #openstack-swift | 19:39 | |
| *** mingyu has quit IRC | 19:39 | |
| *** itlinux has quit IRC | 19:39 | |
| *** openstackgerrit has quit IRC | 19:48 | |
| *** lucasxu has quit IRC | 19:55 | |
| *** skudlik has joined #openstack-swift | 19:56 | |
| *** lucasxu has joined #openstack-swift | 19:56 | |
| *** itlinux has joined #openstack-swift | 19:57 | |
| *** aselius has quit IRC | 20:13 | |
| *** alenavolk has joined #openstack-swift | 20:18 | |
| *** itlinux has quit IRC | 20:18 | |
| acoles | timburke: I think I saw that mattoliverau is offline for a few days, I can cast an eye over your sharding fixups tomorrow, but I'm sure that are great! | 20:24 |
| timburke | ah, right! i forgot | 20:29 |
| *** torgomatic has joined #openstack-swift | 20:30 | |
| *** ChanServ sets mode: +v torgomatic | 20:30 | |
| *** oshritf_ has joined #openstack-swift | 20:31 | |
| *** mingyu has joined #openstack-swift | 20:40 | |
| *** openstackgerrit has joined #openstack-swift | 20:42 | |
| openstackgerrit | Pete Zaitcev proposed openstack/liberasurecode master: Allow to disable optimizations for portability https://review.openstack.org/467761 | 20:42 |
| *** mingyu has quit IRC | 20:44 | |
| *** itlinux has joined #openstack-swift | 20:47 | |
| *** oshritf_ has quit IRC | 20:47 | |
| *** jeffli has joined #openstack-swift | 20:48 | |
| *** lucasxu has quit IRC | 20:49 | |
| *** lucasxu has joined #openstack-swift | 20:53 | |
| kota_ | good morning | 20:56 |
| *** itlinux has quit IRC | 20:57 | |
| notmyname | kota_: good morning. thanks for looking at the per-policy config patch yesterday. your help on that has been very good | 20:57 |
| timburke | good morning kota_! | 20:57 |
| kota_ | notmyname, timburke: hello | 20:58 |
| notmyname | swift team meeting time in #openstack-meeting | 20:59 |
| *** JimCheung has quit IRC | 21:00 | |
| *** JimCheung has joined #openstack-swift | 21:01 | |
| acoles | kota_: I +a's https://review.openstack.org/#/c/465878/, think that is ok | 21:13 |
| patchbot | patch 465878 - swift - Small minor fixes for composite ring functionality | 21:13 |
| kota_ | acoles: thx! | 21:14 |
| timburke | acoles: hah! beat you to it :-) | 21:14 |
| acoles | timburke: think we crossed somewhere mid-atlantic :) | 21:15 |
| *** lucasxu has quit IRC | 21:17 | |
| *** alenavolk has quit IRC | 21:18 | |
| *** NM has quit IRC | 21:29 | |
| openstackgerrit | Merged openstack/swift master: Use setUpModule instead of setup for module level unit test setup https://review.openstack.org/462520 | 21:29 |
| *** oshritf_ has joined #openstack-swift | 21:33 | |
| *** alenavolk has joined #openstack-swift | 21:49 | |
| *** alenavolk has quit IRC | 21:50 | |
| *** alenavolk has joined #openstack-swift | 21:51 | |
| *** alenavolk has quit IRC | 21:53 | |
| onovy | guys, can someone look to https://review.openstack.org/#/c/448244/ and https://review.openstack.org/#/c/391605/ pls? | 22:00 |
| patchbot | patch 448244 - swift - Replace slowdown option with *_per_second option | 22:01 |
| patchbot | patch 391605 - swift - Added rsync_no_change metrics to recon | 22:01 |
| onovy | + notmyname: do you think it's time for stable release? | 22:01 |
| notmyname | onovy: quick question on that | 22:01 |
| notmyname | onovy: the 386 backport for tests is for debian packaging? | 22:01 |
| notmyname | onovy: ie what timburke sort of remembered in his review comment? | 22:01 |
| onovy | yep, replied in gerrit | 22:01 |
| clayg | tdasilva: if it's at midnight for me - then it's like 2am-3am for you right? | 22:01 |
| notmyname | onovy: got it | 22:01 |
| tdasilva | clayg: yep | 22:02 |
| tdasilva | clayg: i'll be reading up the log ;) | 22:02 |
| acoles | tdasilva: lightweight :P | 22:02 |
| tdasilva | lol | 22:02 |
| notmyname | onovy: ok, I'll land those, then, and do a stable release. likely the release will be my tomorrow. depends on the changelog | 22:02 |
| onovy | cool, thanks! | 22:03 |
| *** jeffli has quit IRC | 22:03 | |
| acoles | tdasilva: I'm looking forward to mimicking mattoliverau ... "and now for breakfast" | 22:03 |
| onovy | and what about that two old reviews? :) | 22:03 |
| *** wasmum has joined #openstack-swift | 22:03 | |
| acoles | clayg: so in winter it'll be only 11pm for you | 22:04 |
| timburke | onovy: notmyname: i'd kinda like to sort out why there's the extra ssync in the tests for https://review.openstack.org/#/c/464982/ though :-/ | 22:04 |
| patchbot | patch 464982 - swift (stable/newton) - Do not sync suffixes when remote rejects reconstru... | 22:04 |
| notmyname | here's the meeting time challege: find the row with the least red on it. https://www.timeanddate.com/worldclock/meetingtime.html?month=5&day=31&year=2017&p1=224&p2=43&p3=136&p4=37&p5=204&p6=771&p7=240&p8=248&p9=33&iv=0 | 22:04 |
| tdasilva | acoles: hehe...just not the beach... | 22:04 |
| notmyname | UTC0700 is one of the best times | 22:05 |
| notmyname | red for USA and green everywhere else | 22:05 |
| acoles | tdasilva: I may spread some snd around on my patio, pretend ;) | 22:05 |
| clayg | onovy: patch 448244 looks really good! | 22:05 |
| patchbot | https://review.openstack.org/#/c/448244/ - swift - Replace slowdown option with *_per_second option | 22:05 |
| tdasilva | acoles: heh, get an inflatable shark | 22:06 |
| onovy | clayg: i know! :P | 22:06 |
| onovy | sry, -> bed, gn | 22:09 |
| notmyname | onovy: good night | 22:09 |
| clayg | onovy: I'm still as torn as ever on recon drops for patch 391605 | 22:09 |
| patchbot | https://review.openstack.org/#/c/391605/ - swift - Added rsync_no_change metrics to recon | 22:09 |
| *** oshritf_ has quit IRC | 22:14 | |
| *** vint_bra has quit IRC | 22:18 | |
| *** adriant has joined #openstack-swift | 22:20 | |
| *** mingyu has joined #openstack-swift | 22:28 | |
| *** logan- has joined #openstack-swift | 22:29 | |
| *** joeljwright has quit IRC | 22:31 | |
| *** mingyu has quit IRC | 22:32 | |
| *** itlinux has joined #openstack-swift | 22:45 | |
| *** skudlik has quit IRC | 22:54 | |
| openstackgerrit | Merged openstack/swift master: Small minor fixes for composite ring functionality https://review.openstack.org/465878 | 22:57 |
| *** itlinux has quit IRC | 23:11 | |
| *** klamath has quit IRC | 23:15 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift master: Updated from global requirements https://review.openstack.org/88736 | 23:24 |
| *** kei_yama has joined #openstack-swift | 23:26 | |
| *** alenavolk has joined #openstack-swift | 23:31 | |
| *** logan- has left #openstack-swift | 23:32 | |
| *** chsc has quit IRC | 23:35 | |
| *** alenavolk has quit IRC | 23:38 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: fixup! Replace slowdown option with *_per_second option https://review.openstack.org/467801 | 23:40 |
| *** alenavolk has joined #openstack-swift | 23:50 | |
| *** mingyu has joined #openstack-swift | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!