Friday, 2017-07-07

« Thursday, 2017-07-06 Index Saturday, 2017-07-08 »
notmynameyup. <10m :-)00:00
notmynamenow that all the projects are getting their own docs imported, it would probably be good for all projects to have the same job filters00:01
timburkeman, that sidebar ToC is getting unwieldy00:17
*** Sukhdev has quit IRC00:21
*** mat128 has joined #openstack-swift00:44
tdasilvatimburke: yep00:53
tdasilvaanybody else see probe tests fail?00:53
tdasilvagoing to spin up vagrant-saio to see if I can reproduce there too00:53
timburketdasilva: which one(s)?00:54
*** mat128 has quit IRC00:54
tdasilvatimburke: http://paste.openstack.org/show/614655/00:58
*** tovin07_ has joined #openstack-swift01:05
*** itlinux has joined #openstack-swift01:23
*** itlinux has quit IRC01:23
*** itlinux has joined #openstack-swift01:25
tdasilvaclayg: how does one configure vagrant-saio? I thought it would all be setup once vagrant is done running...01:27
tdasilvaclayg: see this: OSError: [Errno 2] No such file or directory: '/srv/node1'01:28
tdasilvaclayg: got it! was missing the localrc part01:35
*** vint_bra has joined #openstack-swift01:40
kota_good morning01:43
claygI thought defaults should work. If not - that's not good.01:43
* tdasilva is going through a bunch of old bugs and wondering if we should close them....lots of "undecided" "incomplete"/"new" bugs that have not been touched in ages...01:57
openstackgerritTim Burke proposed openstack/swift master: Fix rst for link  https://review.openstack.org/48137702:07
*** mat128 has joined #openstack-swift02:17
openstackgerritCheng Li proposed openstack/swift master: change account_autocreate to default to true  https://review.openstack.org/38594602:21
*** gyee has quit IRC02:23
*** lucasxu has joined #openstack-swift02:53
*** lucasxu has quit IRC02:59
*** vint_bra has quit IRC03:00
*** ukaynar has joined #openstack-swift03:08
*** chlong_ has quit IRC03:08
*** chlong_ has joined #openstack-swift03:08
*** chlong_ has quit IRC03:17
*** chlong_ has joined #openstack-swift03:18
*** psachin has joined #openstack-swift03:27
*** gkadam has joined #openstack-swift03:32
*** mat128 has quit IRC03:39
*** ukaynar has quit IRC03:55
*** JimCheung has quit IRC04:07
*** ukaynar has joined #openstack-swift04:10
*** ukaynar has quit IRC04:26
*** ukaynar has joined #openstack-swift04:26
*** ukaynar has quit IRC04:26
*** links has joined #openstack-swift04:28
*** winggundamth has joined #openstack-swift04:34
*** abhitechie has joined #openstack-swift04:49
*** rcernin has joined #openstack-swift05:46
*** ChubYann has quit IRC05:50
*** cschwede_ has joined #openstack-swift06:06
*** skudlik has joined #openstack-swift06:25
*** gabor_antal_ has joined #openstack-swift06:35
*** timburke has quit IRC06:37
*** rickflare has quit IRC06:37
*** htruta has quit IRC06:38
*** charz has quit IRC06:38
*** gabor_antal has quit IRC06:38
*** notmyname has quit IRC06:38
*** notmyname has joined #openstack-swift06:38
*** ChanServ sets mode: +v notmyname06:38
*** timburke has joined #openstack-swift06:38
*** ChanServ sets mode: +v timburke06:38
*** charz has joined #openstack-swift06:39
*** rickflare has joined #openstack-swift06:40
*** htruta has joined #openstack-swift06:42
*** SkyRocknRoll has quit IRC06:57
openstackgerritAlex Szarka proposed openstack/swift master: Use more specific asserts in tests  https://review.openstack.org/47646406:59
*** itlinux has quit IRC07:16
*** tesseract has joined #openstack-swift07:17
mahatic_I think this patch 385946 needs to have a consensus as it intends to change default for account_autocreate07:25
patchbothttps://review.openstack.org/#/c/385946/ - swift - change account_autocreate to default to true07:25
mahatic_Is cheng li in the channel?07:26
*** abhitechie has quit IRC07:42
acolesgood morning07:54
*** abhitechie has joined #openstack-swift07:59
*** NM has joined #openstack-swift08:30
*** NM has quit IRC08:38
*** psachin_ has joined #openstack-swift09:04
*** psachin_ has quit IRC09:05
mahatic_acoles: good morning09:06
openstackgerritMerged openstack/swift master: docs migration from openstack-manuals  https://review.openstack.org/48122109:07
acoleswow that was quick ^^09:08
acolesmahatic_: hi09:08
mahatic_acoles: merged in 13mins!09:09
*** abhinavtechie has joined #openstack-swift09:22
*** sanchitmalhotra has quit IRC09:23
*** sanchitmalhotra has joined #openstack-swift09:24
*** gkadam_ has joined #openstack-swift09:25
*** gkadam is now known as Guest2449409:25
*** gkadam_ is now known as gkadam09:25
*** abhitechie has quit IRC09:25
acolesmahatic_: right! so now we need to start fixing the docs :)09:29
*** NM has joined #openstack-swift09:30
mahatic_acoles: yup :)09:31
kota_acoles, notmyname, clayg, tdasilva: https://gist.github.com/bloodeagle40234/bf344d9936efe18b5ffbfc79e898636a09:42
kota_current status for isa-l issue09:42
kota_it looks like m = 4, k>=22 cases causes decode error09:43
acoleskota_: m is num parity?09:43
kota_that gist includes my greedy test script and the output09:43
kota_acoles: yes09:43
kota_the output for isa_l_rs_vand09:43
kota_now I'm running on isa_l_rs_cauchy09:44
acoleskota_: fantastic work, thank you09:44
kota_but it didn't finish yet09:44
acoleskota_: did you find *any* errors for cauchy yet?09:44
kota_no errors right now09:45
kota_it has been running since around noon09:45
kota_4-5 hours running09:45
kota_still not yet finished09:46
acolesok09:46
kota_i think, if no errors, it takes longer time than the case if error found09:46
kota_because my test script stops to search other decoding combination for the k, m if an error found09:47
kota_i'll run also other ec engine like jerasure/liberasurecode_rs_vand in the next week but i don't think significant bad pattern found09:48
*** openstackgerrit has quit IRC09:48
kota_maybe09:48
acoleskota_: right, makes sense, so search is truncated on failure but exhaustive for good schemes, so cauchy will hopefully take longer to test exhaustively09:48
kota_acoles: yes, that's what I meant.09:48
kota_hmm, it looks good time to get back home and have a dinner09:58
kota_see you guys09:59
acoleskota_: have a good weekend, thanks again for that work09:59
kota_acoles: you too for having a good weekend ;-)09:59
*** gkadam_ has joined #openstack-swift10:03
*** gkadam has quit IRC10:05
*** tovin07_ has quit IRC10:07
*** gkadam__ has joined #openstack-swift10:10
*** gkadam_ has quit IRC10:12
*** kei_yama has quit IRC11:00
tdasilvagood morning11:19
*** psachin has quit IRC11:36
*** winggundamth has quit IRC11:39
*** zhurong has joined #openstack-swift11:41
*** zhurong has quit IRC11:46
*** pcaruana has joined #openstack-swift11:49
*** NM has quit IRC11:52
*** mat128 has joined #openstack-swift12:09
*** catintheroof has joined #openstack-swift12:40
*** chlong_ has quit IRC12:43
*** gkadam__ has quit IRC12:58
*** ukaynar has joined #openstack-swift12:59
*** lucasxu has joined #openstack-swift12:59
*** links has quit IRC13:06
*** abhinavtechie has quit IRC13:10
*** NM has joined #openstack-swift13:17
*** zhurong has joined #openstack-swift13:23
*** Dinesh_Bhor has quit IRC13:32
*** mat128 has quit IRC13:46
*** ukaynar has quit IRC13:47
*** mat128 has joined #openstack-swift13:49
*** oshritf has joined #openstack-swift14:07
*** chlong_ has joined #openstack-swift14:08
*** oshritf has quit IRC14:11
*** vint_bra has joined #openstack-swift14:17
*** itlinux has joined #openstack-swift14:19
*** zhurong has quit IRC14:26
*** klamath has joined #openstack-swift14:30
*** klamath has quit IRC14:30
*** klamath has joined #openstack-swift14:31
*** skudlik has quit IRC14:41
*** saint_ has joined #openstack-swift14:50
*** catintheroof has quit IRC14:52
*** catintheroof has joined #openstack-swift14:53
*** itlinux has quit IRC14:53
*** catinthe_ has joined #openstack-swift14:56
*** ukaynar has joined #openstack-swift14:56
*** catintheroof has quit IRC14:59
*** lucasxu has quit IRC15:04
*** rcernin has quit IRC15:06
*** catintheroof has joined #openstack-swift15:06
*** catinthe_ has quit IRC15:09
*** openstackgerrit has joined #openstack-swift15:16
openstackgerritMathias Bjoerkqvist proposed openstack/swift master: Retrieve encryption root secret from Barbican  https://review.openstack.org/36487815:16
*** abhitechie has joined #openstack-swift15:20
*** catinthe_ has joined #openstack-swift15:35
*** vint_bra has quit IRC15:36
*** catintheroof has quit IRC15:36
tdasilvamathiasb: hi, what's that link again for the ansible/vagrant script you had to setup barbican?15:42
*** itlinux has joined #openstack-swift15:49
*** ChubYann has joined #openstack-swift15:51
*** gyee has joined #openstack-swift15:58
*** abhinavtechie has joined #openstack-swift15:58
*** abhitechie has quit IRC15:58
*** oshritf has joined #openstack-swift16:00
notmynamegood morning16:04
*** aagrawal has joined #openstack-swift16:04
*** aagrawal has quit IRC16:05
*** abhinavtechie has quit IRC16:08
*** abhitechie has joined #openstack-swift16:12
notmynamekota_: thanks for the EC testing16:13
*** oshritf has quit IRC16:16
*** abhitechie has quit IRC16:16
*** abhitechie has joined #openstack-swift16:17
*** catinthe_ has quit IRC16:19
*** catintheroof has joined #openstack-swift16:19
*** saint_ has quit IRC16:23
*** chsc has joined #openstack-swift16:27
*** chsc has joined #openstack-swift16:27
*** vint_bra has joined #openstack-swift16:28
*** abhinavtechie has joined #openstack-swift16:28
*** abhitechie has quit IRC16:28
*** JimCheung has joined #openstack-swift16:33
*** abhitechie has joined #openstack-swift16:42
*** abhinavtechie has quit IRC16:43
*** psachin has joined #openstack-swift16:47
*** abhitechie has quit IRC16:54
*** psachin has quit IRC16:57
claygtimburke: good one!  I forgot about account autocreate!17:00
timburkeclayg: more broadly, it just seems really silly that we have to separately pass around the method when we already have a request object which necessarily has that information17:02
timburkei'd *love* it if we had a single source of truth17:02
claygyeah... there's *tons* of tests that do BS like req = Request.blank(); controller.PUT(req)17:03
claygi couldn't zero into to a case where it mattered - but I was mostly looking around COPY/POST for objects - which now that they use subrequests from middleware don't have this problem as much17:03
*** hieulq has quit IRC17:12
*** cschwede_ has quit IRC17:18
*** links has joined #openstack-swift17:25
*** hieulq has joined #openstack-swift17:26
*** links has quit IRC17:29
*** links has joined #openstack-swift17:30
openstackgerritTim Burke proposed openstack/swift master: First pass at admin-guide cleanup  https://review.openstack.org/48170417:32
notmynametimburke: done!!17:34
notmynametimburke: I'm working on moving the install guide (then the api docs) next. there is a *lot* of cleanup there too, but I'm only working on the mechanics right now17:34
timburkeyup17:35
notmynamethis whole things sortof resurrects patch 38683417:35
patchbothttps://review.openstack.org/#/c/386834/ - swift - Unify docs trees17:35
notmynameit's a slightly different structure. my old patch made unique doc trees buildable with separate tox jobs that had a common conf.py17:36
*** links has quit IRC17:36
timburkethis probably oughta be a topic for denver: figure out what we want our docs to become, and maybe even do some communal roasts of what we've got now and trying to figure out what we should have instead17:37
notmynameseems that the new world order for docs has a single docs tree where subsets may be individually buildable, but they are also all build via the existing "dev docs"17:37
notmynameyeah, the existing dev docs are sortof not addressed at all in the docs migration work17:37
timburkenotmyname: what i really want to know is how cross-linking should happen now17:37
notmynameand I've already added a topic for the PTG about docs (but maybe lets not wait until then?)17:37
notmynamewhat do you mean by cross linking? what to what?17:38
timburkelike https://review.openstack.org/#/c/481221/2/doc/source/admin/objectstorage-components.rst@11117:39
patchbotpatch 481221 - swift - docs migration from openstack-manuals (MERGED)17:39
timburkeseems better to just link to the ring overview doc17:39
*** ukaynar has quit IRC17:41
notmynameoh, right. there's a ton of duplication now, i suspect. like that17:41
notmynamethe good news with the single source tree (as opposed to my old multi-source tree) is that cross-linking should be very easy17:42
notmynameseems like we'll end up with one and only one set of docs, parts of which are exposed in different places17:42
notmynamehowever, I suspect we'll have everything exposed via swift.openstack.org17:42
notmynameie our "dev docs" will end up being a superset of everything17:43
notmynamewhich IMO is a pretty good thing that we should all be very happy with17:43
*** tonanhngo has joined #openstack-swift17:44
timburkeright, but i've got a suspicion that "parts of which are exposed in different places" will have some impact on the cross-linking -- if everything under doc/source/admin gets published somewhere without the rest of the dev docs, what happens to that link?17:46
notmynamedepends on how the redirect works17:47
notmynameif it "just" ends up being an actual redirect, then it should all work17:47
notmynamehmm... I'm about to propose a patch I think will break. or maybe not. but that's why I'm proposing it. because it will be closer to what I actually want, long-term (ie docs job definitions in tox.ini)17:48
notmynameugh, that's not going to work, or at least be fast18:01
notmynameI was hoping to see a simple way to unify https://github.com/openstack/swift/tree/master/api-ref/source and https://github.com/openstack/swift/tree/master/doc/source/api18:01
notmynamethe former is mostly good. the latter is mostly bad (but not completely, so I don't want to just delete it)18:01
*** bit_lySLH2uSZHed has joined #openstack-swift18:02
notmynameanyone know where the install guide docs job is defined? surely it's looking in a well-known place for the resulting html so that it can be copied to wherever it's published18:04
notmyname*surely*18:04
notmynameor maybe not and it discovers where the output is18:05
*** bit_lySLH2uSZHed has left #openstack-swift18:05
notmynameanyway, if I'm moving the install guide from a top-level dir into the docs tree (yay!) then I need to change the tox.ini job definition for it18:06
notmynamethe current way is to publish the install guide docs inside the existing top-level directory, but that's going away, so what else needs to change (the job definition?) so that the docs can be picked up?18:06
openstackgerritJohn Dickinson proposed openstack/swift master: moved install guide  https://review.openstack.org/48171418:07
openstackgerritJohn Dickinson proposed openstack/swift master: updated docs job in tox.ini  https://review.openstack.org/48171518:07
notmynameI predict that this whole docs migration stuff will likely be the final death of swift.openstack.org. and I'll be very sad about that, and disagree strongly, etc etc etc18:10
notmynamebut I've been told in the past that wanting to keep that is wrong and totally ad odds with everything right and good in the world18:11
*** tesseract has quit IRC18:21
*** itlinux has quit IRC18:27
*** klamath has quit IRC18:32
*** chlong_ has quit IRC18:40
*** itlinux has joined #openstack-swift18:47
openstackgerritThiago da Silva proposed openstack/swift master: Fix swiftdir option and usage of storage policy aliases  https://review.openstack.org/34469318:55
*** ukaynar has joined #openstack-swift18:55
mathiasbtdasilva: the links to the Vagrantfile and the related shell setup script are here https://etherpad.openstack.org/p/swift-kms-keymaster-setup18:57
*** Sukhdev has joined #openstack-swift18:58
tdasilvamathiasb: thanks, I had tried spinning up the vm a while back, ran into some error, but didn't stop to debug, will try again...18:59
mathiasbtdasilva: I ran into some problems with the latest stable branches on devstack a month or two ago (maybe due to some backports?), but then I changed to master and it seemed to work again19:04
tdasilvatimburke, clayg: I was playing around with the probe tests I saw yesterday and they seem to be due to a timing issue. It typically happens in a test when restarting a service (e.g., proxy) and then sending a req, then I get a connection refused error19:05
tdasilvamathiasb: oh i see, DEVSTACK_BRANCH=master, right?19:05
timburke...that would explain some of the failures i've seen in the community cluster lately...19:05
claygmaybe something to do with SO_REUSEPORT stuff rledisez fixed?19:06
claygIIRC the change was "eventlet will use SO_REUSEPORT now so try and bind sooner"19:06
claygmaybe manager.restart has a slightly different behavior wrt shutdown and restart or something... idk19:07
mathiasbtdasilva: right19:08
tdasilvaclayg: yeah, I tried doing manager.restart(wait=True) for some tests, and that seemed to fix it19:08
tdasilvabut then as I ran probetests again, other tests would now fail :(19:08
tdasilvaactually that's incorrect...manager.restart(wait=True) did not work for me, I had to put a sleep after manager.restart()19:09
tdasilvathe partpower_increase tests fail pretty consistently for me19:10
tdasilvaboth on ansible_saio and vagrant_saio19:10
tdasilvaclayg: are you referring to this change? https://github.com/openstack/swift/commit/5d7b0d1edb279cb5713365690fce856a94d3b28919:11
tdasilvamathiasb: running now19:11
claygyeah, that's the one I was thinking of - of course it might be unrelated19:11
*** Renich has joined #openstack-swift19:12
*** acoles has left #openstack-swift19:12
*** ukaynar has quit IRC19:20
*** klamath has joined #openstack-swift19:21
*** ukaynar has joined #openstack-swift19:21
*** ukaynar has quit IRC19:53
*** catinthe_ has joined #openstack-swift19:56
*** catintheroof has quit IRC19:59
openstackgerritPete Zaitcev proposed openstack/swift master: PUT+POST and its development test  https://review.openstack.org/42791120:23
notmynameoh! I'm surprised the docs jobs passed for patch 48171420:25
patchbothttps://review.openstack.org/#/c/481714/ - swift - moved install guide20:25
tdasilvaclayg: reverting rledisez change fixed all probe tests for me20:26
claygtdasilva: oh no!  rledisez20:26
claygit's a good change (in principle) so we'll have to figure out what to do!20:26
*** ukaynar has joined #openstack-swift20:27
*** klamath has quit IRC20:36
*** klamath has joined #openstack-swift20:37
*** ukaynar has quit IRC20:38
*** ukaynar has joined #openstack-swift20:39
*** catinthe_ has quit IRC20:39
openstackgerritMerged openstack/swift master: First pass at admin-guide cleanup  https://review.openstack.org/48170420:42
notmynametimburke: I think maybe we should add in the manuals-jobs to our gate if docs are touched. not sure. maybe we should have in the past but it doesn't matter now? maybe it's more important now that we're taking on more of the docs?20:52
timburkenotmyname: what triggers that job currently?20:53
*** ukaynar has quit IRC20:53
notmynamedunno. I'm only looking at the job definitions. don't know where it's called from20:53
notmynameI'd guess from openstack-manuals project changes20:53
timburkethat's always been one of the problems i've had with project-config...20:53
notmynameya, looks like just some docs projects20:54
notmynameaccording to global find20:54
notmynamehmm... but now I'm looking at swift's definition in projects.yaml, but I don't see how you updated it to only run docs jobs when only docs are changed20:55
notmyname...have I mentioned today how much I love yaml?20:56
notmynameYAML IS TERRIBLE! THERE ARE 63 WAYS TO FORMAT A MULTILINE STRING IN YAML!!!20:56
claygnotmyname: you say that a lot... do you have a source reference?20:57
notmynamehttps://stackoverflow.com/questions/3790454/in-yaml-how-do-i-break-a-string-over-multiple-lines20:58
notmynametl;dr 2 block styles, each with 2 possible block chomping indicators (or none), and with 9 possible indentation indicators (or none), 1 plain style and 2 quoted styles: 2 x (2 + 1) x (9 + 1) + 1 + 2 = 6321:00
*** youplaboum has joined #openstack-swift21:02
timburkenotmyname: i'm not so sure "only run docs jobs when docs are changed" is a good idea -- isn't a lot of the middleware documentation generated out of the swift/ tree, for example?21:04
notmynametimburke: the docs jobs (ie manuals) are only the install guide and api guide. those shouldn't be affected by middleware (or other code modules), right?21:06
*** mat128 has quit IRC21:08
timburkeah, i thought you were including jobs like gate-swift-docs-ubuntu-xenial under that term. so...probably safe? on the other hand, is there significant harm?21:08
timburke(in letting them run on all commits, that is)21:08
notmynameharm to not running manuals jobs?21:08
notmynameoh, no there's no harm AFAIK21:09
notmynamewhere'd you define the limits?21:09
notmynamewhatever it is that you did that allows the gate to go fast if only the docs are touched21:10
notmyname"hey google! where's that thing timburke did that time to make things faster?"21:10
timburkesee https://github.com/openstack-infra/project-config/commit/633b4c69321:12
notmynameah I see21:13
notmynameit's the opposite of what I assumed21:13
timburkethere *is* a way to do what you're thinking of... i know the reno jobs do that... i forget where it's defined though21:14
timburkehttps://github.com/openstack-infra/project-config/blob/master/zuul/layout.yaml#L3298-L3303 maybe? idk21:17
*** ChubYann has quit IRC21:25
*** itlinux has quit IRC21:26
*** catintheroof has joined #openstack-swift21:33
*** NM has quit IRC21:34
*** ChubYann has joined #openstack-swift21:37
*** openstackgerrit has quit IRC21:47
*** tonanhngo has quit IRC21:51
*** openstackstatus has quit IRC21:56
*** openstack has joined #openstack-swift21:59
*** openstackstatus has joined #openstack-swift21:59
*** ChanServ sets mode: +v openstackstatus21:59
*** openstackgerrit has joined #openstack-swift22:16
openstackgerritClay Gerrard proposed openstack/swift master: Don't rehash primaries in reconstructor handoffs_only mode  https://review.openstack.org/42840822:16
clayg^ that one is a pretty good risk/value optimization for handoffs_only mode in the reconstructor - makes handoffs_only on the reconsturctor work more like the replicator - I have some empirical evidence it's effective/safe as well22:18
claygI think rledisez does something similar too22:18
claygkota_: timburke: I think you both have looked at various earlier revisions of that one before22:19
*** klamath has quit IRC22:21
kota_clayg: ack, i added me to the reviewer at gerrit22:22
notmynameFYI if you're planning on going to the sydney summit, australia requires a visa (at least for US citizens--maybe for your country too). many countries can get an electronic visa at https://www.eta.immi.gov.au/ETAS3/etas22:39
notmynamethe ETA is good for 1 year22:39
claygtdasilva: I just did a freshy-freshy build of vsaio with default config/no localrc22:41
claygit worked?22:41
claygtdasilva: do you remember when/where/how/why you saw `OSError: [Errno 2] No such file or directory: '/srv/node1'`22:42
claygor what specifically you changed/added/configured in localrc that made it go away?22:42
clayg... strange ...22:42
*** catintheroof has quit IRC22:42
*** vint_bra has quit IRC23:16
*** SkyRocknRoll has joined #openstack-swift23:49
« Thursday, 2017-07-06 Index Saturday, 2017-07-08 »

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!