Friday, 2015-03-06

*** achanda has quit IRC00:04
openstackgerritJoshua Harlow proposed openstack/taskflow: Avoid checking for overlaps as we already ensure this  https://review.openstack.org/16195300:06
*** jungleboyj has quit IRC00:11
*** jungleboyj has joined #openstack-oslo00:11
openstackgerritJoshua Harlow proposed openstack/taskflow: Avoid checking for overlaps as we already ensure this  https://review.openstack.org/16195300:12
*** zzzeek has joined #openstack-oslo00:20
*** jungleboyj has quit IRC00:23
*** stevemar has quit IRC00:27
*** crc32 has quit IRC00:29
*** liusheng has quit IRC00:29
*** liusheng has joined #openstack-oslo00:29
*** daniel3_ has quit IRC00:33
openstackgerritJoshua Harlow proposed openstack/taskflow: Specialize checking for overlaps  https://review.openstack.org/16195300:34
*** achanda has joined #openstack-oslo00:41
*** jaosorior has quit IRC00:42
*** dims has quit IRC00:44
*** achanda has quit IRC00:46
openstackgerritJoshua Harlow proposed openstack/taskflow: Specialize checking for overlaps  https://review.openstack.org/16195300:55
*** tsekiyam_ has joined #openstack-oslo01:16
*** mtanino has quit IRC01:19
*** tsekiyama has quit IRC01:20
*** tsekiyam_ has quit IRC01:20
openstackgerritJoshua Harlow proposed openstack/taskflow: Add FETCHING notification state (not persisted)  https://review.openstack.org/16185801:20
*** zzzeek has quit IRC01:23
*** achanda has joined #openstack-oslo01:46
*** ChuckC has quit IRC02:04
*** ChuckC has joined #openstack-oslo02:05
*** stevemar has joined #openstack-oslo02:08
*** flwang has quit IRC02:10
*** flwang1 has joined #openstack-oslo02:16
*** crc32 has joined #openstack-oslo02:17
*** tsekiyama has joined #openstack-oslo02:27
*** tsekiyama has quit IRC02:31
*** zzzeek has joined #openstack-oslo02:37
*** zzzeek_ has joined #openstack-oslo02:42
*** zzzeek has quit IRC02:42
*** zzzeek_ is now known as zzzeek02:42
*** zzzeek has quit IRC02:47
*** zzzeek has joined #openstack-oslo02:48
*** zzzeek has quit IRC02:48
*** ChuckC_ has joined #openstack-oslo02:56
*** ChuckC has quit IRC02:59
*** yamahata has quit IRC03:12
*** ChuckC_ is now known as ChuckC03:27
*** david-lyle is now known as david-lyle_afk03:44
*** harlowja is now known as harlowja_away03:57
*** achanda has quit IRC03:57
*** jungleboyj has joined #openstack-oslo03:59
*** takedakn has joined #openstack-oslo04:04
*** amrith is now known as _amrith_04:04
*** takedakn has quit IRC04:08
*** mriedem_away has quit IRC04:10
*** mriedem has joined #openstack-oslo04:11
*** amotoki has joined #openstack-oslo04:16
*** _amrith_ is now known as amrith04:16
*** alexpilotti has quit IRC04:22
*** achanda has joined #openstack-oslo04:38
*** YorikSar has quit IRC04:49
*** exploreshaifali has joined #openstack-oslo04:50
*** achanda has quit IRC04:54
*** pmathews has joined #openstack-oslo04:58
*** exploreshaifali has quit IRC05:14
*** pmathews has quit IRC05:23
*** exploreshaifali has joined #openstack-oslo05:40
*** achanda has joined #openstack-oslo05:56
*** YorikSar has joined #openstack-oslo05:58
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.utils: Imported Translations from Transifex  https://review.openstack.org/16064906:10
*** yamahata has joined #openstack-oslo07:00
*** yamahata has quit IRC07:23
*** achanda has quit IRC07:44
*** achanda has joined #openstack-oslo07:44
*** mriedem has quit IRC07:47
*** mriedem has joined #openstack-oslo07:47
*** achanda has quit IRC07:48
*** liusheng has quit IRC07:50
*** liusheng has joined #openstack-oslo07:51
*** achanda has joined #openstack-oslo08:05
openstackgerritJoshua Harlow proposed openstack/taskflow: Add locks in needed places to avoid corruption  https://review.openstack.org/16205108:11
*** stevemar has quit IRC08:42
*** inc0 has joined #openstack-oslo08:43
*** crc32 has quit IRC08:46
*** dtantsur is now known as dtantsur|bbl08:48
*** jgrimm is now known as zz_jgrimm08:49
*** achanda has quit IRC08:54
*** liusheng has quit IRC08:55
*** liusheng has joined #openstack-oslo08:56
*** andreykurilin_ has joined #openstack-oslo08:56
*** achanda has joined #openstack-oslo08:57
*** stannie has joined #openstack-oslo09:01
*** ajo has joined #openstack-oslo09:05
*** achanda has quit IRC09:12
*** achanda has joined #openstack-oslo09:13
*** achanda has quit IRC09:14
*** achanda has joined #openstack-oslo09:14
*** achanda has quit IRC09:19
*** dulek has joined #openstack-oslo09:30
*** ihrachyshka has joined #openstack-oslo09:33
*** ihrachyshka has quit IRC09:57
*** e0ne has joined #openstack-oslo10:09
*** saikrishna has joined #openstack-oslo10:16
*** andreykurilin_ has quit IRC10:18
*** andreykurilin__ has joined #openstack-oslo10:18
*** e0ne is now known as e0ne_10:22
*** liusheng has quit IRC10:23
*** liusheng has joined #openstack-oslo10:24
*** e0ne_ has quit IRC10:33
*** andreykurilin__ has quit IRC10:38
*** e0ne has joined #openstack-oslo10:39
*** liusheng has quit IRC10:41
*** liusheng has joined #openstack-oslo10:42
*** inc0_ has joined #openstack-oslo10:55
*** inc0 has quit IRC10:56
*** dims has joined #openstack-oslo11:00
*** dulek_ has joined #openstack-oslo11:02
*** dulek has quit IRC11:05
*** e0ne is now known as e0ne_11:07
*** e0ne_ has quit IRC11:18
*** inc0_ has quit IRC11:19
*** dtantsur|bbl is now known as dtantsur11:39
*** inc0 has joined #openstack-oslo11:41
*** e0ne has joined #openstack-oslo11:47
*** flaper87 has quit IRC11:47
*** flaper87 has joined #openstack-oslo11:47
*** inc0 has quit IRC11:55
*** dulek___ has joined #openstack-oslo11:59
*** dulek_ has quit IRC12:01
*** inc0 has joined #openstack-oslo12:11
*** alexpilotti has joined #openstack-oslo12:13
*** inc0 has quit IRC12:15
*** inc0 has joined #openstack-oslo12:16
*** dulek___ has quit IRC12:18
*** dulek___ has joined #openstack-oslo12:18
*** cdent has joined #openstack-oslo12:36
*** cdent has quit IRC12:44
*** saikrishna has quit IRC12:48
*** YorikSar has quit IRC12:56
*** YorikSar has joined #openstack-oslo12:57
*** cdent has joined #openstack-oslo13:03
*** e0ne is now known as e0ne_13:12
*** e0ne_ is now known as e0ne13:22
*** kgiusti has joined #openstack-oslo13:23
*** amrith is now known as _amrith_13:25
*** YorikSar has quit IRC13:28
*** YorikSar has joined #openstack-oslo13:28
*** stpierre has joined #openstack-oslo13:35
*** AAzza_afk has quit IRC13:37
*** AAzza_afk has joined #openstack-oslo13:38
*** gordc has joined #openstack-oslo13:43
*** inc0 has quit IRC13:45
*** sigmavirus24_awa is now known as sigmavirus2413:54
*** exploreshaifali has quit IRC13:55
*** miqui has joined #openstack-oslo13:56
*** ihrachyshka has joined #openstack-oslo14:01
*** dulek___ has quit IRC14:04
*** dulek___ has joined #openstack-oslo14:04
*** zz_jgrimm is now known as jgrimm14:07
*** salv-orl_ has joined #openstack-oslo14:09
*** salv-orlando has quit IRC14:09
dhellmanngood morning14:17
haypomorning dhellmann14:17
dimshi dhellmann haypo14:17
*** saikrishna has joined #openstack-oslo14:17
dhellmannI'm planning to work on code reviews today. Do any of you have anything suggestions for starting places? Anything critical that needs a review?14:19
*** _amrith_ is now known as amrith14:23
*** jecarey has quit IRC14:27
*** dansmith is now known as superdan14:30
*** dims has quit IRC14:37
*** dims has joined #openstack-oslo14:38
*** jungleboyj has quit IRC14:40
dimsdhellmann: couple of low hanging ones from me https://review.openstack.org/#/c/144839/ https://review.openstack.org/#/c/159646/14:44
*** mtanino has joined #openstack-oslo14:51
*** boris-42 has quit IRC14:52
*** e0ne is now known as e0ne_14:55
dhellmanndims: +214:56
dimsdhellmann: requirements bump for oslo.context please https://review.openstack.org/#/c/161868/14:56
dimsthanks dhellmann14:56
*** exploreshaifali has joined #openstack-oslo14:56
*** amotoki has quit IRC14:57
dhellmanndims: +214:59
openstackgerritMerged openstack/oslo-incubator: Handle non-json http exceptions better  https://review.openstack.org/14483915:00
*** e0ne_ has quit IRC15:01
*** dims is now known as dimsum__15:07
*** dulek___ has quit IRC15:16
*** dulek has joined #openstack-oslo15:16
*** zzzeek has joined #openstack-oslo15:20
*** jecarey has joined #openstack-oslo15:32
*** e0ne has joined #openstack-oslo15:32
*** stannie has quit IRC15:36
*** jungleboyj has joined #openstack-oslo15:40
*** ChuckC_ has joined #openstack-oslo15:41
*** ChuckC has quit IRC15:42
*** ChuckC_ has quit IRC15:46
dhellmannjd__: do you have a minute to chat about https://review.openstack.org/#/c/136842/11 ?15:57
jd__dhellmann: sure!15:57
dhellmannso, what's the story with RFCSysLogHandler? it's broken under eventlet?15:58
*** stevemar has joined #openstack-oslo16:01
dhellmannjd__: how does the syslog module know how to communicate with the server? I'm still worried that this represents an incompatible change. :-/16:01
jd__dhellmann: syslog() is a system call16:03
jd__dhellmann: so that logger uses that system call to log16:03
dhellmannjd__: right, but how does that call know where the daemon is?16:03
jd__the problem with the current RFCSysLogHandler is that it's not portable (e.g. /dev/log does not exist everywhere)16:03
jd__dhellmann: you mean the syslog daemon?16:03
dhellmannIOW, if we accept this and suddenly start using the system call is that going to require a change to the syslog configuration in order to work?16:03
jd__dhellmann: no16:03
dhellmannhow do you know that?16:03
jd__the syscall logs to the kernel16:03
dhellmannah16:04
*** dulek has quit IRC16:04
jd__your syslog daemon read the kernel and log to a file16:04
jd__(I say syscall but that's more likely libc call, it may be not that deep)16:05
dhellmannok, I wasn't sure if it was like that or just an abstraction that had to be configured in a different way16:05
dhellmannjd__: how about the eventlet question from greg?16:05
jd__no, the point is not to have to configure it :-)16:05
jd__let me find it16:06
dhellmannright, that makes sense, it just wasn't clear we were actually getting that :-)16:06
jd__ah right, so this function won't be monkey patched by eventlet so it'll "block"16:06
jd__that's why I responded it was good because it'll remove races conditions from our code :->16:07
jd__I don't think syslog() is that blocking as it's not really I/O16:07
*** e0ne is now known as e0ne_16:07
dhellmannthat may depend on the implementation16:08
jd__indeed16:08
jd__I think it's a good chance unless I miss some regression or cases16:08
jd__s/change/change/16:08
jd__I tested locally on my Darwin system and it makes oslo.log work16:09
dhellmannyeah, I'm trying to think through implications for real systems, though16:09
jd__I don't think syslog is enabled by default on devstack nor devstack-gate16:09
jd__so we won't see it there16:09
dhellmannright, this is a production thing -- no one uses it on os x16:09
jd__if someone wants to deploy Gnocchi on BSD it might not work16:10
jd__(devil advocate mode)16:10
*** e0ne_ is now known as e0ne16:10
*** YorikSar has quit IRC16:10
dhellmannthere are ways to make that possible without also causing issues for larger scale users16:10
dhellmannwe could add a new option, for example (as much as I hate having that)16:11
jd__point being?16:11
*** ChuckC_ has joined #openstack-oslo16:12
dhellmannrun for a cycle with the new option, then deprecate it and make it the default behavior to give us feedback16:12
jd__not changing a code because we are afraid does not sound like a good "behaviour"16:12
dhellmannyou're on the receiving end of the hate mail16:12
dhellmann*not16:12
jd__I mean we are supposed to own and masterize this code :-)16:12
jd__haha16:12
dhellmanntrue, but we have real users so sometimes we have to make changes more slowly to allow for adjustments16:13
jd__actually I'm not even sure a lot of people are using use_syslog=True already16:13
dhellmannaside from the eventlet question, I'm happy with this change now16:13
dhellmannyeah, not knowing who uses this feature is part of why changing it concerns me16:13
dhellmannI don't assume that the number is small just because I don't know them16:14
*** AAzza_afk is now known as AAzza16:14
*** jaypipes has joined #openstack-oslo16:14
*** jaypipes is now known as leakypipes16:14
jd__dhellmann: so what should we do?16:16
* dhellmann thinks16:16
dhellmannjd__: a "comprehensive" approach would be to rename --use-syslog to something like --use-syslog-direct and then add a new --use-syslog-syscall (maybe with better names)16:18
dhellmannthen we keep the old behavior for the renamed option and add the new option for using the system call16:18
dhellmannthen next cycle we can deprecate --use-syslog-direct for removal16:18
dhellmannwe could also do something smart like check to see if /dev/log exists, and if it does use the old logger and if it doesn't use the new logger16:20
dhellmann"smart"16:20
sigmavirus24So I think this might affect stackforge/os-ansible-deployment but I'm checking16:20
*** daniel3_ has joined #openstack-oslo16:20
*** sabari has quit IRC16:20
*** stannie has joined #openstack-oslo16:21
dhellmannjd__: the 2nd option is less invasive, and gives us osx support while retaining eventlet support when /dev/log is present16:21
dhellmannsigmavirus24: how so?16:21
sigmavirus24dhellmann: it's the current community project for deploying openstack via ansible (& lxc) but I was wrong.16:22
sigmavirus24It doesn't affect it (because the project configures logging for all services it deploys)16:22
*** sabari has joined #openstack-oslo16:22
sigmavirus24none of the services are configured with use_syslog = True so it shouldn't affect that project16:22
jd__dhellmann: 2nd option sounds like "chicken" :)16:23
jd__I think we should simplify our live and merge this patch as is16:24
* dhellmann jd__ you have NO idea how much shit I get when you guys break things16:24
jd__bah there's a revert button no? :)16:24
sigmavirus24jd__: I think it's a good change if we could make it less problematic16:24
sigmavirus24Also not breaking things randomly is a good overall goal16:25
jd__dhellmann: so how about we turn it on by default and add a config to make it back to the old if necessary? And we deprecate this option right away.16:25
dhellmannjd__: that could result in breaking CD systems16:25
jd__dhellmann: err if that's the case then my patch is wrong16:25
jd__dhellmann: the feature should be 100% the same AFAICT16:26
dhellmannright, and that's what I can't tell because I don't have anywhere I can test this at scale16:26
jd__so that's our problem16:26
dhellmannwell, except you acknowledge that eventlet won't be patched so performance might suffer16:26
dhellmann*won't be monkeypatching16:26
jd__probably 0.1% slower but might fix a few bugs :p16:27
dhellmannwhat I would *rather* do is not have --use-syslog at all, and let deployers use a logging config file to set that up16:27
jd__so what's the procedure to change code at scale? we ask people to test and then we merge? :)16:27
jd__ah we can also deprecate use_syslog entirely16:27
dhellmannwhat we've tried to do is put the change in using a backwards-compatible approach, then encourage folks to turn it on to test it and then make it the default over time16:28
*** daniel3_ has quit IRC16:29
jd__do you really think people will test it or is it just a figure to say later "we told you to test it" if it breaks? :)16:29
dhellmanna little of both16:30
sigmavirus24jd__: why not both?16:30
jd__:))16:30
dhellmannit would give us a way to run a gate job with it enabled, though16:30
*** daniel3_ has joined #openstack-oslo16:31
jd__ack, just let me know what I should change in the patch and I will anyway16:31
* dhellmann tries to decide which is least invasivce16:33
*** amotoki has joined #openstack-oslo16:33
dhellmannif the goal is to have a way to use syslog when /dev/log is not present, then let's just put that check in place and switch between the two logger classes16:34
dhellmannwe can revisit deprecation of the direct logger separately that way, and still move ahead adding the support you need16:34
dhellmannjd__: thoughts?16:34
jd__dhellmann: sounds good to me as a first step16:35
dhellmannjd__: ok, good, thanks16:36
*** stpierre has quit IRC16:44
*** stpierre has joined #openstack-oslo16:44
superdandhellmann: thanks for the d-g change, I didn't have that on my radar, sorry16:52
*** mdbooth_ has joined #openstack-oslo16:52
*** yamahata has joined #openstack-oslo16:52
openstackgerritOpenStack Proposal Bot proposed openstack/oslo-incubator: Updated from global requirements  https://review.openstack.org/16158116:52
*** boris-42 has joined #openstack-oslo16:53
dhellmannsuperdan: np, I'll look at the remaining few updates for those sorts of things -- we don't want any projects using the lib until we have the testing in place, and that requires these other updates16:53
superdandhellmann: I thought we were letting heat go ahead after the g-r change?16:55
dhellmannsuperdan: yeah, I forgot about these other things, so we need to try to get them in16:55
dhellmannwe don't want a long period of time without those tests in place16:55
superdandhellmann: these other changes protect us from making another release and breaking them, right? they don't buy us anything for the existing release do they?16:56
superdans/them/heat, etc/16:56
dhellmannsuperdan: right, they protect us from accepting a change that would break if we released it16:56
*** mdbooth has quit IRC16:56
*** mdbooth_ is now known as mdbooth16:56
openstackgerritMerged openstack/oslo-incubator: Remove policy from oslo-incubator  https://review.openstack.org/15281216:57
openstackgerritMerged openstack/oslo-incubator: Prevent update.py from updating policy  https://review.openstack.org/15281316:57
superdandhellmann: right, okay, so if heat is okay with the current release they can still iterate, right? We can just not merge anything else in the lib until we know the tests are in place, yes?16:57
dhellmannsuperdan: yeah16:58
superdandhellmann: okay cool16:58
openstackgerritMerged openstack/oslo.messaging: Fix matchmaker_redis ack_alive fails with KeyError  https://review.openstack.org/15567317:00
openstackgerritMerged openstack/oslo-specs: Dropping graduate-oslo-reports  https://review.openstack.org/15873617:01
*** tsekiyama has joined #openstack-oslo17:01
dhellmannharlowja_away: it looks like https://blueprints.launchpad.net/oslo-incubator/+spec/adopt-debtcollector is done, is that right?17:02
openstackgerritOpenStack Proposal Bot proposed openstack/oslo-incubator: Updated from global requirements  https://review.openstack.org/16158117:04
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.messaging: Updated from global requirements  https://review.openstack.org/15778317:04
dhellmannstevemar, ayoung, morganfainberg : it looks like https://blueprints.launchpad.net/oslo-incubator/+spec/graduate-policy is done,  is that right?17:05
stevemardhellmann, was just gonna update that17:05
dhellmannstevemar: I'll let you do the honors :-)17:06
stevemarwoo hoo17:06
stevemardhellmann, i think the honors are yours, i am not an oslo-incubator driver, or the original assignee :)17:07
dhellmannstevemar: ah, ok, well, done17:07
stevemar\o/17:07
*** e0ne is now known as e0ne_17:16
*** e0ne_ is now known as e0ne17:16
*** cdent has quit IRC17:22
*** ajo has quit IRC17:22
*** YorikSar has joined #openstack-oslo17:25
*** e0ne is now known as e0ne_17:27
*** dtantsur is now known as dtantsur|afk17:30
*** dtantsur|afk is now known as dtantsur|pto17:30
*** dulek has joined #openstack-oslo17:31
*** e0ne_ has quit IRC17:32
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.log: Updated from global requirements  https://review.openstack.org/16158317:38
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.middleware: Updated from global requirements  https://review.openstack.org/16158417:38
openstackgerritOpenStack Proposal Bot proposed openstack/oslo.versionedobjects: Updated from global requirements  https://review.openstack.org/16091817:38
*** pblaho has quit IRC17:38
*** dulek has quit IRC17:39
dhellmanndimsum__, bnemec: does this failure mean anything to either of you? http://logs.openstack.org/49/160649/4/check/gate-oslo.utils-python34/2ef951f/console.html17:40
bnemecdhellmann: Hmm, looks like a problem in excutils.  We did just merge a change to that.17:42
*** david-lyle_afk is now known as david-lyle17:43
dhellmannbnemec: it's working for me here locally :-/17:45
bnemecOh, it literally got 'expected string or buffer' as the exception message.17:46
bnemecWeird.17:46
dhellmannyeah, something is getting an  object, throwing an error with that message, and then that isn't matching the expected error17:46
dhellmannit's python 3, so maybe something got a byte string instead of text?17:46
*** zzzeek has quit IRC17:49
dimsum__looking17:49
*** stevemar has quit IRC17:50
bnemecThe change that just merged altered the mocking around that test, maybe py3 is unhappy with one of the new mocked values or something.17:50
*** stevemar has joined #openstack-oslo17:50
dhellmannah, yeah, if it got a mock that would do it -- but I wonder why the tests worked on the patch that was merged17:51
dimsum__bnemec: harlowja_away's commit? https://github.com/openstack/oslo.utils/commit/2e6517151e2c85e89e8ba3055f7ab9428f49570417:52
bnemecdimsum__: Yeah, that's the one I'm thinking of.17:52
*** bnemec is now known as beekneemech17:52
* dimsum__ be right back, need some lunch17:52
openstackgerritMerged openstack/oslo.messaging: Use import of zmq package for test skip  https://review.openstack.org/15442617:53
*** pmathews has joined #openstack-oslo17:53
openstackgerritMerged openstack/oslo.vmware: Move missing tests to new oslo_vmware location  https://review.openstack.org/16106217:54
beekneemechStrange, it passes under py34 locally for me as well. :-/17:59
beekneemechThat log is really confusing.  Supposedly this error happened 14 seconds before the test commands were actually executed?18:04
*** yamahata has quit IRC18:07
openstackgerritMerged openstack/oslo-incubator: Updated from global requirements  https://review.openstack.org/16158118:16
*** YorikSar has quit IRC18:18
*** morganfainberg is now known as needscoffeebadly18:19
stevemarbeekneemech, ha! i just noticed your fun friday name, good one.18:20
*** pmathews has quit IRC18:20
beekneemechstevemar: <3 casual nick friday :-)18:20
*** pmathews1 has joined #openstack-oslo18:20
needscoffeebadlydhellmann: yay!18:21
needscoffeebadlyOslo.policy yay18:21
*** needscoffeebadly is now known as CaptainMorgan18:22
*** e0ne has joined #openstack-oslo18:22
stevemarCaptainMorgan, where you been!18:23
*** achanda has joined #openstack-oslo18:23
stevemarlate to the game as usual18:23
stevemari sent that email yesterday18:23
CaptainMorganstevemar: shhhhhh18:23
*** e0ne has quit IRC18:24
*** jogo is now known as flashgordon18:27
openstackgerritMerged openstack/cliff: Document print_help_if_requested method  https://review.openstack.org/15980818:28
*** salv-orl_ has quit IRC18:28
*** YorikSar has joined #openstack-oslo18:29
openstackgerritMerged openstack/oslo.db: Updated from global requirements  https://review.openstack.org/16139418:30
*** yamahata has joined #openstack-oslo18:31
mriedemrpodolyaka: dimsum__: dhellmann: is this code just used in testing? https://review.openstack.org/#/c/120870/18:42
mriedemi'm a bit concerned about requireing testresources in non-test-requirements18:42
dimsum__mriedem: good point18:47
mriedemb/c a distro that ships oslo.db 1.5.0+ now needs to ship testresources as a runtime dep18:47
mriedemwhich is wonky18:47
mriedemthe only problem i see is with oslo_db.sqlalchemy.utils.is_backend_avail imports provision18:49
mriedemwhich imports testresources now18:49
mriedemthat method is deprecated though18:49
mriedemlooks like is_backend_avail is only used in tests though18:49
mriedemi'm gonna open a bug18:50
*** stevemar is now known as stevedore18:50
*** achanda has quit IRC18:52
mriedemhttps://bugs.launchpad.net/oslo.db/+bug/142923318:52
openstackLaunchpad bug 1429233 in oslo.db "oslo.db now requires testresources at runtime" [Undecided,New]18:52
*** achanda has joined #openstack-oslo18:57
openstackgerritMatt Riedemann proposed openstack/oslo.db: Move testresources/testscenarios into test-requirements  https://review.openstack.org/16227218:58
mriedemdimsum__: ^18:58
mriedemrpodolyaka: ^18:58
mriedemmaybe that provision module needs to be moved under tests18:58
dimsum__mriedem: ack taking a look19:01
*** zzzeek has joined #openstack-oslo19:05
*** daniel3_ has quit IRC19:05
*** daniel3__ has joined #openstack-oslo19:05
*** saikrishna has quit IRC19:08
*** pmathews1 has quit IRC19:09
openstackgerritDavanum Srinivas (dims) proposed openstack/oslo.db: Move testresources/testscenarios into test-requirements  https://review.openstack.org/16227619:20
dimsum__mriedem: counter review ^^^19:20
mriedemtricky19:22
mriedemdimsum__: add the bug number in the commit message19:22
mriedemplz19:22
dimsum__ok19:22
openstackgerritDavanum Srinivas (dims) proposed openstack/oslo.db: Move testresources/testscenarios into test-requirements  https://review.openstack.org/16227619:23
dimsum__mriedem: added you as well19:24
mriedem+119:27
openstackgerritDean Troyer proposed openstack/cliff: Pass user command text to the Command object  https://review.openstack.org/14738119:27
*** daniel3__ has quit IRC19:35
*** daniel3_ has joined #openstack-oslo19:36
*** ajo has joined #openstack-oslo19:37
*** YorikSar has quit IRC19:37
dhellmanndimsum__, mriedem: https://review.openstack.org/#/c/158787/19:42
dhellmannoh, that's not the one I was thinking of19:43
dhellmanndimsum__, mriedem : https://review.openstack.org/#/c/157135/1/specs/policy/external-library-requirements.rst,cm19:44
openstackgerritDoug Hellmann proposed openstack/oslo-specs: Add general policy on test tooling  https://review.openstack.org/15878719:45
mriedemdhellmann: unless i'm missing something, that seems to backup that test libs shouldn't be imported to requirements.txt19:46
dhellmannI swear we wrote this specific case down somewhere when we decided to do it this way.19:48
dhellmannmriedem: line 44 of that diff19:48
mriedemso oslo_db.sqlalchemy.provision is meant for public consumption?19:49
*** CaptainMorgan is now known as morganfainberg19:50
mriedembut the public consumption would be tests in other projects, right?19:50
dhellmannmriedem: I don't know off the top of my head, but the module name makes me think so19:50
dhellmannyes19:50
mriedemgross19:50
mriedemcouldn't we move that into like oslo.db.fixtures or something?19:50
mriedemwe have to package and do legal approvals on everything we ship if the distro (rhel, ubuntu) doesn't provide it,19:50
mriedemso packaging up test libs is not fun19:51
dhellmannmaybe? I don't keep up with every patch in every library. rpodolyaka, viktors|afk, and zzzeek would know better about this specific case19:51
dhellmann"we"?19:51
mriedemibm19:51
mriedemthe product i work on19:51
dhellmannk19:51
zzzeekmriedem: the provision thing is mostly intended for consumption by the sqlalhcemy/test_base.py stuff that is public19:51
*** pmathews has joined #openstack-oslo19:51
dhellmannyeah, we're trying to balance "don't specify too many requirements" and "don't make installing it useless"19:51
zzzeekmriedem: using it directly, id ask why doesnt test_base do what you need19:52
*** salv-orlando has joined #openstack-oslo19:52
mriedemdo what i need?19:52
dhellmannzzzeek: https://review.openstack.org/15878719:52
mriedemi'm not looking to integrate with this19:52
dhellmannzzzeek: we're going to have to stop expecting projects to use test base classes :-/19:52
mriedemas a packager, i don't want to package testresources/testscenarios for oslo.db 1.5.019:52
zzzeekdhellmann: test_base.py has a fixture in it19:52
zzzeekdhellmann: so, a fixture19:52
zzzeekdhellmann: better ?19:53
dhellmannzzzeek: fixtures are good19:53
zzzeekdhellmann: because, AFAICT, we *are* expecting everyone to be on the testX toolchain, including testresources, testscenarios, etc19:53
zzzeekI of course resisted this a lot19:53
mriedemthat's totally fine19:53
mriedemas long as it's only required in test-requirements for a project,19:54
mriedemor if the project is a test-only project itself19:54
mriedeme.g. oslo.db.fixtures19:54
zzzeekmriedem: but that’s where its at.  the thigns oslo.db needs to do, are in testresources and testscenrarios19:54
zzzeekmriedem: meaning what, spin it out into a new project?  sure19:54
mriedemyes19:54
zzzeekmriedem: that woudl be great19:54
zzzeekmriedem: make it so19:54
mriedemalright19:54
dhellmannwait, what?19:54
zzzeekah19:54
zzzeekhah19:54
*** pmathews has quit IRC19:55
zzzeekdhellmann: mriedem is proposing that we make a new project that featrues testing fixtrues to use with relational database setups19:55
dhellmannzzzeek: we can expect them to use fixtures, but mixing test base classes started causing initialization order issues in nova.19:55
dhellmannnot this cycle19:55
dhellmannfeature freeze is next thursday19:55
zzzeekdhellmann: i didnt say anyhting about any cycles :)19:55
*** pmathews has joined #openstack-oslo19:55
mriedemyeah i'm proposing next week19:56
zzzeekmriedem: oh.  yeah probably not19:56
mriedembut right now, for packaging, i have to sed out those requirements from requirements.txt19:56
zzzeekmriedem: this woudl be a big serious rip the thing apart at the seams kind of change19:56
mriedemso stevedore doesnt puke on startup when they aren't there19:56
zzzeekmriedem: well then that means essentially revert the whole feature19:56
dhellmannthe problem with that approach though is we end up having twice as many libraries19:56
stevedoremriedem, i'm always here19:56
dhellmannmaybe it's better just to go back to saying you have to put those test dependencies in your own requirements list19:56
mriedemstevedore: puking?19:56
zzzeekmriedem: im not really sure how the gerrit got +2’ed if now we have to revert it?   is there a regression ?19:56
stevedoremriedem, all the time19:56
openstackgerritDan Krause proposed openstack/taskflow: Persistence backend refactor  https://review.openstack.org/16088919:57
mriedemzzzeek: i'm not saying revert19:57
mriedemzzzeek: https://bugs.launchpad.net/oslo.db/+bug/142923319:57
openstackLaunchpad bug 1429233 in oslo.db "oslo.db now requires testresources/testscenarios at runtime" [Undecided,In progress] - Assigned to Matt Riedemann (mriedem)19:57
zzzeekmriedem: testresources is a dependencty for provision.py which is a dependency for oslo.db’s tests to run19:57
mriedemzzzeek: yes, i know19:57
mriedemi'm saying, things required for tests should be in test-reqiurements19:57
*** pmathews has quit IRC19:57
mriedemthe problem is provision is a public thing for other projects, like nova19:57
mriedemso provision is not in the /tests dir19:58
zzzeekmriedem:  that dependency is requied for *everyones* tests that are currently based on oslo_db.sqlalchemy.test_base19:58
mriedemtherefore testresources is not in test-requirements19:58
mriedemsure19:58
zzzeekmriedem: so…if oyu need it out next week, we just revert19:58
zzzeekmriedem: its not essential right now, nobody’s really using the elaborate version of the feature19:58
zzzeekdhellmann: FYI this is why i *Really really really* did not want to use all the testresources stuff19:58
zzzeekmriedem: i had this all working without any of this stuff.  lifeless -1’ed it19:59
mriedemlike i said, i believe my workaround for now is to sed out the test* entries from requirements.txt before building a package19:59
zzzeekmriedem: go complain to him :)19:59
zzzeekmriedem: i’m not up to speed on our specific building systems, so its btween you and dhellmann for now :)20:00
* zzzeek doing actual employer work! woop20:00
mriedembut now these test libs are in everyone's non-test import path if they are using oslo.db, which is everyone20:00
dhellmannzzzeek: this problem isn't specific to testresources, it's also just an issue with test dependencies in general20:00
mriedemand that grosses me out20:00
zzzeeki just get annoyed when, somethign grosses *me* out, nobody cares, i get rolled over, then four months later, people see it all committed and theyre like WHAT ?!20:01
zzzeek:)20:01
mriedemping me next time :)20:01
mriedemwe can be grossed out together20:01
*** amrith is now known as _amrith_20:04
*** mriedem has quit IRC20:05
*** salv-orlando has quit IRC20:11
*** exploreshaifali has quit IRC20:13
*** pmathews has joined #openstack-oslo20:13
*** mriedem has joined #openstack-oslo20:13
mriedemsorry, system got borked20:14
mriedemzzzeek: dhellmann: should i start a ML thread on this to talk about options?20:14
openstackgerritDan Krause proposed openstack/taskflow: Persistence backend refactor  https://review.openstack.org/16088920:15
dhellmannmriedem, zzzeek : we also have that spec and can hold the discussion there20:24
*** pmathews has quit IRC20:25
mriedemdhellmann: posted to ML and linked to the spec since i already had it written up20:28
dhellmannmriedem: ok20:28
mriedemnow apparently i was supposed to have started on a thing that i got an urgent email about at 8am which i didn't see until 2pm :)20:29
mriedemoops20:29
dhellmannoops20:29
*** YorikSar has joined #openstack-oslo20:31
*** tedross has joined #openstack-oslo20:35
*** tedross has left #openstack-oslo20:36
*** pmathews has joined #openstack-oslo20:36
*** achanda has quit IRC20:37
*** pmathews has quit IRC20:38
lifelessmriedem: whats the dependency issue ?20:39
mriedemhttp://lists.openstack.org/pipermail/openstack-dev/2015-March/058460.html20:39
lifelessok20:40
lifelessso the pip upstream folk actually don't like what we do with requirements20:41
lifelesstheir opinion is that install_requires should only hard-require things needed to to ensure import works, and that actual 'works with X' are an install choice for the end product - e.g. our requirements.txt should not fold into install_requires, and distributors should choose their own requirements.txt based on their needs20:42
lifelessbasically you're getting bleed through from that20:42
lifelessthis is conceptually no different than e.g. not having hard deps on sqlite/pymysql/pypgsql etc20:42
lifelessI'll follow up on the list on Monday20:42
*** crc32 has joined #openstack-oslo20:43
mriedemwell, more and more optional backend deps are moving to test-requirements20:43
mriedemlike oslo.vmware in nova is now in test-requirements, or should be soon20:43
lifelessbut my tl;dr is that testscenarios would still be test-requirements for oslo.db20:43
lifelessnot requirements20:43
mriedemthat was the change i proposed20:43
lifelessmakes sense to me20:43
mriedemhttps://review.openstack.org/#/c/162272/20:43
lifelessthe separate package thing allows crisper requirements and I support that too20:44
mriedemdimsum__: i've restored https://review.openstack.org/#/c/162272/20:44
mriedemsince it doesn't sound like that provision module should move20:44
mriedemunless it's to a separate project20:44
mriedemin L20:44
*** exploreshaifali has joined #openstack-oslo20:45
*** pmathews has joined #openstack-oslo20:46
*** pmathews has quit IRC20:47
*** pmathews has joined #openstack-oslo20:50
*** stannie has quit IRC20:54
dhellmannmriedem: I thought you said something else in oslo.db imported the provision module, though?20:56
dhellmannif that's true, we have to fix that before we remove the dependency20:56
mriedemdhellmann: oslo_db.sqlalchemy.utils.is_backend_available, which itself is deprecated and only imported into other test modules20:56
dhellmannmriedem: if it's part of the public API of the library, then even though it is deprecated we can't be sure that no apps are using it20:57
dhellmannwhich means we need to make that work without those test dependencies before we can remove them20:57
*** achanda has joined #openstack-oslo21:00
*** dimsum__ has quit IRC21:01
*** pmathews has quit IRC21:05
*** pmathews has joined #openstack-oslo21:06
*** ajo has quit IRC21:11
*** kgiusti has left #openstack-oslo21:15
*** ihrachyshka has quit IRC21:26
*** stpierre has quit IRC21:27
*** harlowja_away has quit IRC21:34
*** jungleboyj has quit IRC21:34
*** salv-orlando has joined #openstack-oslo21:36
*** dimsum__ has joined #openstack-oslo21:41
jecareydhellmann, I'd like to talk to you about this spec when you have a few min https://review.openstack.org/#/c/121459/21:42
*** amotoki has quit IRC21:48
*** exploreshaifali has quit IRC21:56
*** prad has joined #openstack-oslo21:56
dhellmannjecarey: sure, what's up?21:59
openstackgerritMerged openstack/oslo.messaging: Updated from global requirements  https://review.openstack.org/15778322:02
jecareydhellmann, I think I helped let this one slip between the cracks.  Wondering if any chance for it in Kilo now.22:03
dhellmannjecarey: it's pretty late :-/  is there code?22:04
jecareydhellmann, yes, he has the code.22:04
jecareyhttps://review.openstack.org/#/c/134850/22:05
*** YorikSar has quit IRC22:05
jecareydhellmann, still need to figure out right answer to questoin of how to gather the messages.22:06
*** jgrimm is now known as zz_jgrimm22:06
dhellmannthat and, do we need to support contextual forms of all of the existing log-level translator functions?22:06
dhellmannand plurals?22:06
jecareyI'd asked it in an inline comment and he restated with more detail in his comments.22:06
dhellmannI wonder if we should really be extending the API of the existing method with some optional keyword args, rather than making new methods22:07
jecareyI don't think we need to support it in the logs.  I think this is rare.22:07
dhellmannok22:07
dhellmanndo we have an example of an actual case where we want to do this in an application?22:08
jecareyI've seen some cases in Russian that I helped debug.22:09
jecareyThat was actually in the horizon stuff and was a company translation.22:09
dhellmannok. I don't know enough about this part of the library to know how common it really is, so I wanted to make sure we weren't trying for completeness if we didn't need the feature22:09
jecareyIt is rare ... but when it happens there isn't enough information to do the right translation without it.22:10
dhellmannwhat's an example of a "contextual marker"? is it just a message to the translator, or is it something the user sees somewhere?22:11
openstackgerritMerged openstack/oslo.middleware: Updated from global requirements  https://review.openstack.org/16158422:12
jecareymonth = pgettext("month name", "May")22:13
*** pblaho has joined #openstack-oslo22:13
jecareySo it provides a hint to the translator when the text is ambigious22:13
*** pblaho has quit IRC22:13
dhellmannok, I see22:14
dhellmannjecarey: let me spend a little time reading the spec and code. Our cut-off is supposed to be next Thursday, so this is pretty late, but since it all exists we should look at it and see what we can do.22:15
jecareydhellmann, I appreciate it.   Can I ask a procedural question?22:16
*** dimsum__ has quit IRC22:16
openstackgerritMerged openstack/oslo.log: Updated from global requirements  https://review.openstack.org/16158322:16
*** dimsum__ has joined #openstack-oslo22:17
dhellmannjecarey: of course22:17
jecareyWhat's my responsiblity as the -1er?   I kind of expected the author to bring it into the IRC or the meeting.22:17
jecareyI just happened to notice this hadn't moved.22:18
dhellmannjecarey: it's their work, they should follow up22:18
dhellmannI don't remember seeing a ML thread22:19
jecareyI don't either and it doesn't reference one.22:20
*** jungleboyj has joined #openstack-oslo22:20
jecareydhellmann, thanks for the help/guidance.22:21
*** dimsum__ has quit IRC22:21
*** e0ne has joined #openstack-oslo22:22
*** dimsum__ has joined #openstack-oslo22:23
dhellmannjecarey: any time! and don't worry too much about it, we have things fall off of the radar every release, that's why we just keep going :-)22:24
*** andreykurilin_ has joined #openstack-oslo22:24
*** exploreshaifali has joined #openstack-oslo22:26
flashgordonzzzeek: got a db question for you on #openstack-nova if you are around22:27
zzzeekflashgordon: ya22:27
*** zzzeek has quit IRC22:32
*** YorikSar has joined #openstack-oslo22:37
*** dimsum__ has quit IRC22:43
*** dimsum__ has joined #openstack-oslo22:43
*** mriedem has quit IRC22:47
*** dimsum__ has quit IRC22:48
*** salv-orlando has quit IRC22:53
*** ajo has joined #openstack-oslo22:53
*** achanda_ has joined #openstack-oslo23:03
*** achanda has quit IRC23:03
*** ajo has quit IRC23:03
*** andreykurilin_ has quit IRC23:04
*** cdent has joined #openstack-oslo23:04
*** achanda_ has quit IRC23:07
*** daniel3_ has quit IRC23:09
*** achanda has joined #openstack-oslo23:10
*** gordc has quit IRC23:10
*** cdent has quit IRC23:12
*** exploreshaifali has quit IRC23:12
*** crc32 has quit IRC23:13
*** prad has quit IRC23:16
*** e0ne has quit IRC23:22
*** ChuckC_ has quit IRC23:25
*** stevedore has quit IRC23:31
*** pmathews has quit IRC23:40
*** pmathews has joined #openstack-oslo23:41
*** ChuckC_ has joined #openstack-oslo23:43
*** ChuckC_ is now known as ChuckC23:44
*** dimsum__ has joined #openstack-oslo23:59

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!