| *** yamahata has joined #openstack-oslo | 00:42 | |
| openstackgerrit | Joshua Harlow proposed a change to openstack/oslo.messaging: Start adding a futures executor based executor https://review.openstack.org/70914 | 01:05 |
|---|---|---|
| openstackgerrit | Joshua Harlow proposed a change to openstack/oslo.messaging: Start adding a futures executor based executor https://review.openstack.org/70914 | 01:06 |
| *** rpodolyaka has quit IRC | 01:23 | |
| *** rpodolyaka has joined #openstack-oslo | 01:25 | |
| *** rpodolyaka has quit IRC | 01:52 | |
| *** rpodolyaka has joined #openstack-oslo | 01:52 | |
| openstackgerrit | Joshua Harlow proposed a change to openstack/oslo.messaging: Start adding a futures executor based executor https://review.openstack.org/70914 | 01:53 |
| *** lbragstad has joined #openstack-oslo | 02:00 | |
| *** rpodolyaka has quit IRC | 02:10 | |
| *** yamahata has quit IRC | 02:11 | |
| *** rpodolyaka has joined #openstack-oslo | 02:30 | |
| *** zhiyan_ is now known as zhiyan | 02:51 | |
| *** harlowja is now known as harlowja_away | 02:54 | |
| *** rpodolyaka has quit IRC | 02:58 | |
| *** harlowja_away is now known as harlowja | 03:06 | |
| *** wingwj has quit IRC | 03:24 | |
| *** wingwj has joined #openstack-oslo | 03:25 | |
| openstackgerrit | Luis A. Garcia proposed a change to openstack/oslo-incubator: Add default user_identity to logging record https://review.openstack.org/79533 | 04:08 |
| *** yamahata has joined #openstack-oslo | 04:12 | |
| *** rpodolyaka has joined #openstack-oslo | 05:54 | |
| *** harlowja is now known as harlowja_away | 06:08 | |
| openstackgerrit | Joshua Harlow proposed a change to openstack/taskflow: Lock test tweaks https://review.openstack.org/79401 | 06:53 |
| *** rpodolyaka has quit IRC | 07:01 | |
| *** rpodolyaka has joined #openstack-oslo | 07:01 | |
| *** rpodolyaka has joined #openstack-oslo | 07:01 | |
| *** rpodolyaka has quit IRC | 07:02 | |
| *** rpodolyaka has joined #openstack-oslo | 07:07 | |
| *** rpodolyaka has quit IRC | 07:17 | |
| openstackgerrit | Mehdi Abaakouk proposed a change to openstack/oslo-incubator: Raise RuntimeError of unloadable library https://review.openstack.org/77934 | 07:40 |
| openstackgerrit | Mehdi Abaakouk proposed a change to openstack/oslo-incubator: Raise RuntimeError of unloadable library https://review.openstack.org/77934 | 07:45 |
| openstackgerrit | Mehdi Abaakouk proposed a change to openstack/oslo-incubator: Re-raise exception of unloadable library https://review.openstack.org/77934 | 07:46 |
| *** flaper87|afk is now known as flaper87 | 07:50 | |
| *** dshulyak has joined #openstack-oslo | 07:54 | |
| *** viktors has joined #openstack-oslo | 08:07 | |
| *** yamahata has quit IRC | 08:14 | |
| *** yamahata has joined #openstack-oslo | 08:30 | |
| openstackgerrit | Mehdi Abaakouk proposed a change to openstack/oslo.messaging: Remove old drivers dead code https://review.openstack.org/78903 | 08:32 |
| openstackgerrit | Ivan Melnikov proposed a change to openstack/taskflow: Test that failures in retry results are saved correctly https://review.openstack.org/79561 | 08:36 |
| *** yamahata has quit IRC | 09:04 | |
| *** gcb has joined #openstack-oslo | 09:06 | |
| openstackgerrit | Ilya Pekelny proposed a change to openstack/oslo-incubator: Fix migrations changing the type of deleted column https://review.openstack.org/72419 | 09:56 |
| *** yamahata has joined #openstack-oslo | 10:09 | |
| ihrachys | dhellmann: up to discuss oslo.messaging notifier interface? | 10:37 |
| flaper87 | dims_: you around? | 10:54 |
| flaper87 | dims_: I've a question about config.generator | 10:54 |
| ihrachys | in oslo-rpc, notifier allowed to specify priority in generic way thru argument to notify(). In oslo.messaging, there are separate functions for error, info... To be able to mimic prev behavior, I can use _notify(), but this is not part of public API. Is it ok? Or should I introduce some hack like a dictionary of priority:function pairs? | 11:04 |
| *** yamahata has quit IRC | 12:09 | |
| dhellmann | ihrachys: don't use the private method -- what's the use case that requires specifying the priority using something other than calling the appropriate method? | 12:09 |
| *** dims_ has quit IRC | 12:12 | |
| ihrachys | dhellmann: well, we already expose priority as conf option, so it's probably for backwards compatibility | 12:14 |
| *** dims_ has joined #openstack-oslo | 12:14 | |
| ihrachys | dhellmann: and some notifications are invoked with this configured priority, whatever it is | 12:15 |
| dhellmann | ihrachys: "we"? | 12:17 |
| ihrachys | dhellmann: eh... as in 'neutron' :) | 12:17 |
| ihrachys | the question comes from my porting effort for neutron | 12:18 |
| dhellmann | where is that in neutron? | 12:18 |
| ihrachys | 1sec | 12:19 |
| ihrachys | dhellmann: https://github.com/openstack/neutron/blob/master/neutron/api/v2/base.py#L343 | 12:20 |
| ihrachys | there are some other instances in the same file | 12:20 |
| ihrachys | also affected: neutron/db/l3_db.py, neutron/services/metering/agents/metering_agent.py, | 12:21 |
| dhellmann | I'm not sure that value should be configurable | 12:21 |
| dhellmann | is that normal to do that? | 12:22 |
| ihrachys | I'm not either, but attempting to remove the option may expose questions unrelated to porting itself, like backwards compatibility concern | 12:22 |
| ihrachys | most notifications in neutron are done without using this conf option | 12:23 |
| dhellmann | part of the point of changing the notifier API was to ensure that the string used as the level was always consistent | 12:23 |
| ihrachys | fair enough. MAYBE this is some error in neutron itself, I may try to ask about that in neutron channel in parallel | 12:24 |
| dhellmann | the application should be controlling the notification "level" for each message, based on the content of the message -- what the message is saying | 12:24 |
| dhellmann | do you know why this was made configurable? | 12:24 |
| ihrachys | no, I don't | 12:25 |
| dhellmann | ok, perhaps we should talk to markmcclain about it | 12:25 |
| dhellmann | IMHO, the config option should be removed, but I would like to understand the history | 12:26 |
| ihrachys | yeah, maybe. OR we may end up with a dictionary of string:function pairs | 12:26 |
| dhellmann | yeah, that would be a way to work around it in neutron | 12:27 |
| ihrachys | the latter option does not touch the current behavior, so the option may be removed separately later. | 12:27 |
| ihrachys | I think I'll go that route now, we'll see later whether we may get rid of it completely later. I'll try to catch mark at #neutron re: the matter. | 12:28 |
| openstackgerrit | Mehdi Abaakouk proposed a change to openstack/oslo.messaging: Implements notification-dispatcher-filter https://review.openstack.org/77886 | 12:35 |
| openstackgerrit | Mehdi Abaakouk proposed a change to openstack/oslo.messaging: Full support of multiple hosts in transport url https://review.openstack.org/78948 | 12:37 |
| openstackgerrit | Mehdi Abaakouk proposed a change to openstack/oslo.messaging: Add transport reconnection retries https://review.openstack.org/75365 | 12:37 |
| dims_ | flaper87, pong | 12:43 |
| flaper87 | dims_: hey, so | 12:46 |
| flaper87 | config generator currently scans files in the project base and tries to find oslo.config options being created there | 12:46 |
| flaper87 | that's done everytime regardless there are explicit entry points specified for the project | 12:47 |
| flaper87 | is there a way to prevent config.generator doing that? | 12:47 |
| flaper87 | The only way I found to do that is to use a tuple instead of a list for the config options | 12:48 |
| flaper87 | (not sure if what I'm saying makes sense) | 12:48 |
| dims_ | flaper87, give me an example please? | 12:48 |
| flaper87 | dims_: I don't want the generator to do this on Marconi's modules: https://github.com/openstack/oslo-incubator/blob/master/openstack/common/config/generator.py#L116 | 12:49 |
| flaper87 | because we already have this: https://github.com/openstack/marconi/blob/master/marconi/queues/bootstrap.py#L53 | 12:49 |
| flaper87 | I thought about overwriting $FILES in the config-generator.rc but that would prevent it from scanning files under openstack/common/ | 12:51 |
| dims_ | flaper87, which piece of code calls_config_options to collect the options? | 12:53 |
| flaper87 | dims_: those are entry_points to use with -l https://github.com/openstack/marconi/blob/master/setup.cfg#L51 | 12:53 |
| flaper87 | since the config generator does this: gihttps://github.com/openstack/oslo-incubator/blob/master/openstack/common/config/generator.py#L116 | 12:54 |
| flaper87 | I can workaround this issue by using `tuples` but that's not nice | 12:55 |
| flaper87 | so, I was wondering if there's a better way to do it | 12:55 |
| flaper87 | and obviously, I think we should improve this somehow | 12:55 |
| dims_ | flaper87, so, the entry points for oslo.config.opts is picked up in lines 111-114, can we save those and skip the same modules in the 115-125 block? | 13:02 |
| dims_ | i.e. when we see "marconi.bootstrap = marconi.queues.bootstrap._config_options" then we should not try to look inside marconi.queues.bootstrap.py? | 13:04 |
| flaper87 | dims_: yeah, pretty much that, or we could simple allow the project to explicitly specify which packages should be scanned | 13:04 |
| flaper87 | i.e: MARCONI_PACKATES_TO_SCAN="marconi/openstack/common/" | 13:05 |
| flaper87 | and then just fils under marconi/openstack/common will be picked | 13:05 |
| flaper87 | the rest goes under MARCONI_GENERATOR_EXTRA_LIBRARIES="..." | 13:05 |
| flaper87 | ideally, all projects should move away from scanning files to just using entry_points for this | 13:06 |
| dims_ | flaper87, my pref is to try to use existing constructs if possible | 13:06 |
| *** gcb has quit IRC | 13:06 | |
| dims_ | flaper87, agree we should move all projects to entry_points | 13:07 |
| flaper87 | what I don't like about the existing constructs is that they support the fact that there are globally defined configuration options, AFAICS | 13:07 |
| openstackgerrit | Stanislav Kudriashev proposed a change to openstack/taskflow: Rework proxy publish functionality https://review.openstack.org/79601 | 13:07 |
| dims_ | flaper87, k. let's check with dhellmann as well, he must have something in mind as well | 13:08 |
| *** mriedem has joined #openstack-oslo | 13:15 | |
| *** lbragstad has quit IRC | 13:18 | |
| *** yamahata has joined #openstack-oslo | 13:26 | |
| *** beav has joined #openstack-oslo | 13:29 | |
| openstackgerrit | Stanislav Kudriashev proposed a change to openstack/taskflow: Use message.requeue instead of message.reject https://review.openstack.org/79607 | 13:32 |
| *** yamahata has quit IRC | 13:41 | |
| *** yamahata has joined #openstack-oslo | 13:41 | |
| dhellmann | dims_, flaper87 : the idea was to use the scanning for apps and the entry points for libraries | 13:44 |
| *** lbragstad has joined #openstack-oslo | 13:44 | |
| dims_ | dhellmann, i am catching flak on #openstack-cinder for this https://bugs.launchpad.net/oslo/+bug/1290861 (see https://bugs.launchpad.net/oslo/+bug/1290503 as well) | 13:45 |
| dims_ | dhellmann, please take a peek for that as well | 13:45 |
| dhellmann | dims_: isn't logging_context_format_string a configuration option for the logging module? | 13:47 |
| flaper87 | dhellmann: mmh, although that makes sense, I still think there should be a way to opt-out from scanning. I may be wrong, though. | 13:47 |
| flaper87 | to be completely honest, this is the first time I mess directly with config.generator | 13:47 |
| dims_ | dhellmann, yes, it is. where can a project specify it is the problem | 13:49 |
| dhellmann | flaper87: does the generator make assumptions about which files to scan, or is it given a list? | 13:50 |
| dhellmann | dims_: set_default()? | 13:50 |
| dhellmann | dims_: as in cfg.CONF.set_default(logging_context_format_string='something without user_identity') | 13:50 |
| dhellmann | dims_: what's the plan to get cinder on the shared context? | 13:50 |
| dims_ | dhellmann, i also see a set_defaults in log.py | 13:51 |
| dhellmann | dims_: ah, yeah, I thought that was there but couldn't find it | 13:51 |
| * dhellmann is still warming up his fingers this morning | 13:51 | |
| dims_ | dhellmann, problem is that cinder.conf.sample would still have the one from the code | 13:52 |
| dims_ | and not the one set using either CONF.set_default or log.set_defaults | 13:53 |
| dhellmann | dims_: true, unless set_defaults() is called in a module during import | 13:54 |
| dims_ | ah | 13:54 |
| dhellmann | dims_: which is ugly, but... | 13:54 |
| dhellmann | dims_: actually, I'm not sure set_defaults() will change the value in the place it needs to be changed, we should try it but it may not work | 13:57 |
| dims_ | dhellmann, too many places to add :( not sure if it's a good idea | 13:58 |
| dhellmann | dims_: hey, what do you know, it does change the option directly | 13:58 |
| dhellmann | too many places? | 13:58 |
| dims_ | cinder/bin has at least 9 scripts i think | 13:58 |
| dhellmann | dims_: why not just add the value to the cinder context, then? | 13:59 |
| dims_ | dhellmann, right after say "LOG = logging.getLogger(__name__)"? | 14:01 |
| dhellmann | dims_: no, in to_dict() add a key with the name 'user_identity' built like we do in the oslo context -- just copy that little bit of the code over | 14:01 |
| dhellmann | I guess it needs an instance_uuid too | 14:01 |
| dhellmann | not instance_uuid, just instance | 14:02 |
| dhellmann | ugh | 14:02 |
| dhellmann | we should make that "instance" key more generic to let the other projects override it | 14:03 |
| dhellmann | I'd like to get to the point where cinder subclasses our context and adds to it, instead of reimplementing it | 14:03 |
| dims_ | dhellmann, nova just went the other way since we told them we are getting rid of our context no? | 14:04 |
| * dhellmann didn't know we were getting rid of our context | 14:04 | |
| dhellmann | when did we decide that? | 14:04 |
| dhellmann | if we can't have a common context api it's going to be very difficult to make them all work with the logging module | 14:05 |
| dims_ | == context == | 14:05 |
| dims_ | M: | 14:05 |
| dims_ | S: Obsolete | 14:05 |
| dims_ | F: context.py | 14:05 |
| dhellmann | sigh | 14:05 |
| dhellmann | ok | 14:05 |
| dhellmann | we may regret that, we'll see | 14:06 |
| dhellmann | so what do you think about adding user_identity to the cinder context? | 14:06 |
| dhellmann | it's either that or add calls to change the default context format string to all the required places -- maybe one module that the cinder apps already import? | 14:07 |
| dims_ | dhellmann, i am thinking about the opposite. in oslo's log.py we do this | 14:07 |
| dims_ | extra.setdefault('user_identity', kwargs.pop('user_identity', None)) | 14:07 |
| dims_ | we can just look for user there as well | 14:07 |
| dhellmann | dims_: if we're already doing that, why are we getting a key error? | 14:09 |
| * bnemec wonders that too | 14:09 | |
| dims_ | good point | 14:10 |
| dhellmann | the ColorHandler doesn't call the ContextHandler | 14:11 |
| dhellmann | er, ContextFormatter | 14:11 |
| dhellmann | that's the bug, I think | 14:11 |
| dhellmann | ColorHandler is hard-coded to use the default StreamHandler formatter, but we have a custom formatter | 14:11 |
| dims_ | ha! | 14:12 |
| dhellmann | sorry, power flickered here a few times -- it's a very pretty windless day, so of course | 14:14 |
| dhellmann | dims_, I think we need both fixes -- the context formatter should look for everything that's in the default format string and make sure there's at least an empty string, and then that ColorHandler should use the right formatter | 14:17 |
| dhellmann | dims_, I'll mark https://bugs.launchpad.net/oslo/+bug/1290503 as having high priority for rc1 | 14:18 |
| dims_ | thanks dhellmann please add these notes there as well | 14:18 |
| dhellmann | dims_: are you working on the fix, or should I work something up? | 14:18 |
| dims_ | dhellmann, please go ahead. i will get a chance only late tonight | 14:19 |
| dhellmann | dims_: ok | 14:19 |
| dhellmann | hmm, it's calling its base class method. I wonder why that doesn't call the configured formatter? | 14:20 |
| dhellmann | ok, need more research | 14:20 |
| *** rpodolyaka has joined #openstack-oslo | 14:40 | |
| openstackgerrit | gordon chung proposed a change to openstack/oslo-incubator: sql backend throwing DeprecationWarning: BaseException.message https://review.openstack.org/79630 | 15:04 |
| sileht | jd__, did you see the py33 failure on https://review.openstack.org/#/c/74728/ | 15:07 |
| sileht | jd__, I have the same issue on other oslo-incubator patch | 15:08 |
| sileht | jd__, this is perhaps due to the nose release yesterday | 15:08 |
| sileht | jd__, downgrading fixes the issue :( | 15:18 |
| dhellmann | dims_, bnemec : https://review.openstack.org/#/c/79533/ looks like a reasonable work-around for cinder's keyerror issue, until they can add the value to their context | 15:25 |
| bnemec | dhellmann: It looks like we have precedent for that change, but I feel like it means we're doing something wrong. All of those parameters should have been populated by the ContextAdapter, shouldn't they? | 15:29 |
| dhellmann | bnemec: in this case, the ContextAdapter isn't being used | 15:29 |
| dhellmann | as we move oslo.log to a library, we are going to have to stop relying on the ContextAdapter and move a lot of that stuff into the ContextFormatter | 15:30 |
| dhellmann | oslo.log will set up the formatter, and then all logging calls will have the context info, even from third-party libraries | 15:30 |
| dhellmann | and our libs won't all rely on oslo.log | 15:30 |
| bnemec | dhellmann: Hmm, okay. | 15:31 |
| * bnemec doesn't understand how this works very well | 15:31 | |
| dhellmann | right now we wrap every call that asks for a log handler and add our context adapter to the result | 15:32 |
| dhellmann | that means the context is only available to openstack code | 15:32 |
| dhellmann | that means that when sqlalchemy, for example, logs it does not include request context | 15:32 |
| bnemec | I thought we fixed that. | 15:32 |
| dhellmann | the context formatter fixes it, right? | 15:32 |
| dhellmann | maybe I misunderstood some of the history :-(* | 15:33 |
| bnemec | I remember reviewing a change to pull the context from a thread local store. | 15:33 |
| bnemec | Let me see if I can find it. | 15:33 |
| dhellmann | yeah, that's the context formatter | 15:33 |
| bnemec | Oh, right. It's directly above this change. :-) | 15:33 |
| dhellmann | right, so with the context formatter we shouldn't need the context adapter any more -- except that the formatter isn't doing all of the things the adapter does | 15:34 |
| dhellmann | the section of ContextFormatter.format() that starts with the comment "store request info" needs to have the smarts of ContextAdapter.process() | 15:35 |
| bnemec | Right, okay. So this is just the new way things should work. | 15:35 |
| dhellmann | and we also need to figure out a way to avoid having project-specific keywords in the format strings (like 'instance') | 15:35 |
| dhellmann | bnemec: right | 15:35 |
| bnemec | Yeah, lots of nova-isms left in the logging format. | 15:36 |
| dhellmann | we're moving in the right direction, just not quite there yet | 15:36 |
| ihrachys | dhellmann: discussed with mark, we'll replace conf setting with audit and remove it from conf file | 15:36 |
| bnemec | I know the keystone guys commented on that not too long ago too. | 15:36 |
| dhellmann | ihrachys: sounds good, thanks! | 15:36 |
| dhellmann | bnemec: yeah, it keeps coming up | 15:36 |
| dhellmann | let me make sure we have a bug or something filed | 15:36 |
| dhellmann | blueprint, I guess | 15:36 |
| bnemec | dhellmann: It's kind of a bug too. We probably need blueprints for all of this logging work. | 15:37 |
| bnemec | Obviously I didn't understand it. :-) | 15:37 |
| dhellmann | bnemec: see the dep graph on https://blueprints.launchpad.net/oslo/+spec/graduate-oslo-log | 15:37 |
| *** harlowja_away is now known as harlowja | 15:38 | |
| bnemec | dhellmann: Okay, so we need to add blueprints for "remove the nova-isms" and "remove the need for ContextAdapter" still, right? | 15:38 |
| dhellmann | bnemec: yes, I'm writing up the first of those now | 15:39 |
| ihrachys | flaper87: dhellmann: in neutron, I've seen some agents to use BASE_RPC_API_VERSION instead of RPC_API_VERSION; do you know about the difference between them? It seems BASE_RPC_API_VERSION is not checked by oslo-rpc, while RPC_API_VERSION is. | 15:39 |
| *** rpodolyaka has quit IRC | 15:40 | |
| dhellmann | bnemec: https://blueprints.launchpad.net/oslo/+spec/app-agnostic-logging-parameters | 15:40 |
| openstackgerrit | Joshua Harlow proposed a change to openstack/oslo.messaging: Add a thread + futures executor based executor https://review.openstack.org/70914 | 15:42 |
| dhellmann | ihrachys: I'm afraid I don't know | 15:42 |
| bnemec | dhellmann: Nice. Then related to dims_ comment on https://review.openstack.org/#/c/79533/1 that would just be a workaround until we can set more sane defaults for our logging formats? | 15:42 |
| *** harlowja has quit IRC | 15:43 | |
| dhellmann | bnemec: I thought, incorrectly, that most projects used the oslo request context, and so having user_identity be defined there would make the logging work right on the next sync. I should have checked more closely. | 15:43 |
| bnemec | Or, no, we want everyone to not have to change our logging formats. | 15:44 |
| dhellmann | we have deprecated our request context, but it feels like we at least want something to define what the API for that class should look like -- a simple common base class | 15:44 |
| dhellmann | right, that's the idea -- we use generic terms for parts of the message structure, and the project-specific contexts fill those in with useful info | 15:44 |
| dhellmann | in place of "instance" we might say "object" or something and then in nova it's the instance UUID and in cinder it's the volume UUID | 15:44 |
| bnemec | Okay, think I've got it. Will +A that change. | 15:45 |
| bnemec | Ooh, after Jenkins passes though. :-) | 15:45 |
| dhellmann | yeah, we hit that db race issue | 15:45 |
| bnemec | Yeah, that's going to be a headache for basically the rest of the cycle it sounds like. | 15:46 |
| dhellmann | there's something wrong with only some of the nodes? | 15:46 |
| bnemec | dhellmann: Yeah, according to Roman in https://bugs.launchpad.net/oslo/+bug/1288916 about 40% of the infra nodes don't have the new configuration, and for some reason they can't fix that right now. | 15:47 |
| *** harlowja has joined #openstack-oslo | 15:48 | |
| bnemec | I wonder if we could stick the lock fixture in the db base test class until the problem is fixed. | 15:48 |
| dhellmann | bnemec: that seems like a reasonable workaround, what are the cons? | 15:49 |
| bnemec | dhellmann: It would serialize the db tests, which is not a big deal, and infra told us they'd rather have us do it the other way, but since it's not currently working I'm not too concerned about that either. :-) | 15:51 |
| *** zhiyan is now known as zhiyan_ | 15:51 | |
| dhellmann | if it's only temporary... | 15:51 |
| bnemec | Yeah | 15:51 |
| bnemec | I'll get a change proposed. | 15:51 |
| dhellmann | cool, thanks | 15:51 |
| openstackgerrit | Jenkins proposed a change to openstack/oslo-incubator: Updated from global requirements https://review.openstack.org/77644 | 15:53 |
| openstackgerrit | Jenkins proposed a change to openstack/pycadf: Updated from global requirements https://review.openstack.org/75672 | 15:54 |
| *** rpodolyaka has joined #openstack-oslo | 15:57 | |
| openstackgerrit | Ben Nemec proposed a change to openstack/oslo-incubator: Add lockutils fixtures to OpportunisticTestCase https://review.openstack.org/79655 | 16:05 |
| dhellmann | bnemec: have you ever seen this python 3 error? http://logs.openstack.org/33/79533/1/check/gate-oslo-incubator-python33/79f577e/console.html | 16:05 |
| bnemec | dhellmann: I haven't, but I think someone said it seemed to be related to a new version of nose that just released. | 16:06 |
| bnemec | <sileht> jd__, did you see the py33 failure on https://review.openstack.org/#/c/74728/ | 16:06 |
| bnemec | <sileht> jd__, I have the same issue on other oslo-incubator patch | 16:06 |
| bnemec | <sileht> jd__, this is perhaps due to the nose release yesterday | 16:06 |
| bnemec | <sileht> jd__, downgrading fixes the issue :( | 16:06 |
| bnemec | dhellmann: ^ | 16:06 |
| dhellmann | bnemec: aha | 16:07 |
| jd__ | let me sleep | 16:07 |
| dhellmann | jd__: no sleep for the wicked | 16:07 |
| bnemec | jd__: Never! | 16:07 |
| jd__ | :) | 16:07 |
| jd__ | wouh, that error sucks | 16:07 |
| dhellmann | so is there a review up somewhere I can +2 to change the version of nose we use? | 16:07 |
| jd__ | like downgrading? | 16:08 |
| dhellmann | yeah, this is the gate so we have to specify the upper bound in our requirements list | 16:08 |
| dhellmann | global-requirements.txt just says "nose" | 16:09 |
| bnemec | Yeah, we're probably going to have to get that changed before we can fix oslo-incubator. :-/ | 16:09 |
| jd__ | or we go punch the nose upstream authors? | 16:09 |
| dhellmann | jd__: you do that, I'll work on figuring out which version of nose works right | 16:09 |
| jd__ | punch on their nose of course (PUN INTENDED) | 16:09 |
| bnemec | Indeed. :-) | 16:10 |
| jd__ | dhellmann: ack | 16:10 |
| dhellmann | wow, they are replacing self.__class__? | 16:10 |
| dhellmann | wtf? | 16:10 |
| sileht | dhellmann, I bisect nose to find the borken commit | 16:11 |
| dhellmann | sileht: does downgrading to 1.3.0 work, or do we have to go to an older version? | 16:11 |
| sileht | dhellmann, yes | 16:11 |
| jd__ | sileht: did you find the culprit? | 16:12 |
| dhellmann | answered as a mathematician | 16:12 |
| dhellmann | sileht: which? | 16:12 |
| jd__ | I'm opening a bug in their Github | 16:12 |
| sileht | 1.3.0 works | 16:12 |
| dhellmann | sileht: thanks :-) | 16:12 |
| jd__ | sileht: do you have the commit id of the regression or? | 16:12 |
| sileht | jd__, soon | 16:12 |
| * jd__ shakes sileht holding him by the shoulder | 16:12 | |
| jd__ | TELL ME NOW | 16:12 |
| jd__ | sileht: feel free to add https://github.com/nose-devs/nose/issues/780 | 16:13 |
| openstackgerrit | Ben Nemec proposed a change to openstack/oslo-incubator: Add lockutils fixtures to OpportunisticTestCase https://review.openstack.org/79655 | 16:14 |
| sileht | the winner is : https://github.com/nose-devs/nose/commit/c51ff6ffda7e8b36eaa1152090e6238479610d7d | 16:14 |
| *** rpodolyaka has quit IRC | 16:15 | |
| sileht | jd__, dhellmann, beav ^ | 16:15 |
| sileht | bnemec, ^ | 16:15 |
| jd__ | triangulating location of the author | 16:15 |
| jd__ | sending black helicopters… | 16:16 |
| bnemec | The anticipation is killing me. :-) | 16:16 |
| dims_ | lol | 16:17 |
| bnemec | They must not test against py3? im_class went away, didn't it? | 16:17 |
| *** rpodolyaka has joined #openstack-oslo | 16:17 | |
| * dims_ peeks outside the window | 16:17 | |
| dhellmann | sileht, jd__, bnemec : https://review.openstack.org/79661 | 16:18 |
| dhellmann | dims_, did you catch up on our discussion of the logging issue? | 16:19 |
| dims_ | dhellmann, just scrolling back now | 16:20 |
| dhellmann | dims_: tl;dr - the default for user_identity was only in ContextAdapter, not ContextFormatter, so we need luis' patch for now and a cinder patch to add a real value is up for review, too | 16:21 |
| * bnemec lost the window with that logging review. :-( | 16:22 | |
| dhellmann | bnemec; https://review.openstack.org/#/c/79533/ | 16:22 |
| dhellmann | that's currently blocked on the nose issue for python 3 | 16:22 |
| bnemec | Ah, that's right. | 16:22 |
| dims_ | dhellmann, ah ok thanks | 16:23 |
| *** saikrishna_ has joined #openstack-oslo | 16:23 | |
| dhellmann | dims_: we also added https://blueprints.launchpad.net/oslo/+spec/app-agnostic-logging-parameters for juno | 16:24 |
| dims_ | dhellmann, yep saw that | 16:24 |
| dhellmann | bnemec, were you going to write up "remove ContextAdapter" or was I supposed to do that? I can, just don't want a dupe. | 16:24 |
| bnemec | dhellmann: I can do it. | 16:24 |
| bnemec | It'll be a good sanity check that I actually understand what's going on too. :-) | 16:25 |
| dhellmann | bnemec: ok :-) | 16:26 |
| * dhellmann doesn't think sanity is a prerequisite for membership on this team | 16:26 | |
| *** rpodolyaka has quit IRC | 16:27 | |
| bnemec | dhellmann: +1! | 16:28 |
| openstackgerrit | Andrey Kurilin proposed a change to openstack/oslo-incubator: Add an automated naming convention for UNIQUEs https://review.openstack.org/73191 | 16:30 |
| openstackgerrit | James Carey proposed a change to openstack/oslo-incubator: Fix gettextutil.Message handling of deep copy failures https://review.openstack.org/79667 | 16:31 |
| bnemec | Ah, gettextutils. The gift that keeps on giving. :-) | 16:34 |
| bnemec | dhellmann: https://blueprints.launchpad.net/oslo/+spec/remove-context-adapter | 16:39 |
| dhellmann | bnemec: nice, I added a detail to the whiteboard | 16:40 |
| *** rpodolyaka has joined #openstack-oslo | 16:40 | |
| bnemec | dhellmann: Cool | 16:41 |
| openstackgerrit | James Carey proposed a change to openstack/oslo-incubator: Fix gettextutil.Message handling of deep copy failures https://review.openstack.org/79667 | 16:41 |
| dhellmann | we're building up quite a list of blueprints for juno already: https://blueprints.launchpad.net/oslo/juno | 16:41 |
| harlowja | dhellmann are u aware of anyone looking into https://github.com/jsonpickle/jsonpickle and using that for various places instead of jsonutils (jsonpickle seems to have better abilities to deserialize/serialize custom types) | 16:52 |
| dhellmann | harlowja: I'm not. The "reconstituting" part makes me shudder, given the issues with regular pickle, but it may be worth a look | 16:53 |
| harlowja | dhellmann agreed, it does make me shudder to ;) | 16:53 |
| harlowja | jsonpickle takes the jsonutils serializers to the next level (allowing more customization of the serializers/deserializers) | 16:55 |
| openstackgerrit | Roman Podoliaka proposed a change to openstack/oslo-incubator: Add a base test case for DB schema comparison https://review.openstack.org/74081 | 16:57 |
| openstackgerrit | Stanislav Kudriashev proposed a change to openstack/taskflow: Rename remote task to request https://review.openstack.org/79685 | 17:14 |
| *** rpodolyaka has quit IRC | 17:16 | |
| *** arnaud has joined #openstack-oslo | 17:19 | |
| *** rpodolyaka has joined #openstack-oslo | 17:22 | |
| bnemec | dhellmann: Glance actually wants to start using the common context. :-/ | 17:25 |
| bnemec | https://bugs.launchpad.net/glance/+bug/1288846 | 17:25 |
| rpodolyaka | isn't it deprecated in oslo? | 17:26 |
| bnemec | rpodolyaka: Yes, that's the problem. :-) | 17:27 |
| bnemec | It's also being used in oslo db right now, apparently. | 17:27 |
| rpodolyaka | yeah, I talked to viktors this morning about removing it (and we can do that easily as the only function it's passed in isn't used in target projects right now) | 17:28 |
| rpodolyaka | I'm just wondering, why did we deprecate it in the first place? | 17:28 |
| bnemec | rpodolyaka: Can you comment on that bug and let them know that? | 17:29 |
| rpodolyaka | bnemec: sure! | 17:29 |
| bnemec | I don't know. We were just talking about it for logging this morning too, so I'm wondering if we need to revisit that deprecation. | 17:29 |
| rpodolyaka | it's definitely useful for logging, as it's the place we take info about the request being processed from | 17:30 |
| openstackgerrit | Jay Bryant proposed a change to openstack/oslo-incubator: Fix gettextutil.Message handling of deep copy failures https://review.openstack.org/79667 | 17:32 |
| bnemec | rpodolyaka: Thanks! Maybe this was all covered in the deprecation discussion the first time around, but I don't think I was involved with that. | 17:34 |
| openstackgerrit | Jenkins proposed a change to openstack/oslo-incubator: Updated from global requirements https://review.openstack.org/77644 | 17:35 |
| openstackgerrit | Jenkins proposed a change to openstack/oslo.messaging: Updated from global requirements https://review.openstack.org/79691 | 17:35 |
| openstackgerrit | Jenkins proposed a change to openstack/oslo.vmware: Updated from global requirements https://review.openstack.org/79692 | 17:35 |
| openstackgerrit | Jenkins proposed a change to openstack/pycadf: Updated from global requirements https://review.openstack.org/79693 | 17:35 |
| openstackgerrit | Jenkins proposed a change to openstack/taskflow: Updated from global requirements https://review.openstack.org/79698 | 17:37 |
| *** rpodolyaka has quit IRC | 17:38 | |
| *** viktors is now known as viktors_ | 17:39 | |
| bnemec | Hmm, https://review.openstack.org/#/c/77644/7 isn't updating the test-requirements, so it isn't going to fix nose. :-( | 17:41 |
| *** saikrishna_ has quit IRC | 17:47 | |
| dhellmann | bnemec: do you remember why we deprecated it in the first place? | 17:48 |
| dhellmann | bnemec: nm, just caught up on the backlog | 17:48 |
| *** rpodolyaka has joined #openstack-oslo | 17:49 | |
| bnemec | dhellmann: Yeah, I'm assuming that happened at the last summit, which I missed. | 17:49 |
| *** saikrishna_ has joined #openstack-oslo | 17:49 | |
| dhellmann | bnemec: I don't think that requirements change is the latest version, the nose version setting hasn't merged yet (waiting for tests) | 17:49 |
| bnemec | dhellmann: Oh, that makes sense. | 17:50 |
| dhellmann | bnemec: I would like to reconsider dropping the common context, and think about whether it makes sense to keep it as a base class for the other projects to subclass. | 17:50 |
| bnemec | dhellmann: That certainly seems like the right thing to do based on what we know right now. | 17:51 |
| dhellmann | I'll update the GraduationStatus page with a comment to that effect | 17:51 |
| dhellmann | hmm, that page lists it as Orphaned not Obsolete | 17:52 |
| dhellmann | I wonder if it's middleware/context.py that we talked about deleting | 17:52 |
| dhellmann | https://etherpad.openstack.org/p/icehouse-oslo-status | 17:52 |
| openstackgerrit | Julien Vey proposed a change to openstack/oslo-incubator: Don't slugify names that don't exist. https://review.openstack.org/78642 | 17:54 |
| dhellmann | bnemec: https://wiki.openstack.org/wiki/Oslo/GraduationStatus#context | 17:54 |
| bnemec | dhellmann: Hmm, yeah. There may be some conflation of the context middleware and the Context class going on here. | 17:57 |
| dhellmann | bnemec: right | 17:58 |
| openstackgerrit | Roman Podoliaka proposed a change to openstack/oslo-incubator: Add a base test case for DB schema comparison https://review.openstack.org/74081 | 18:13 |
| *** mriedem has quit IRC | 18:16 | |
| *** rpodolyaka has quit IRC | 18:24 | |
| *** rpodolyaka has joined #openstack-oslo | 18:27 | |
| *** mriedem has joined #openstack-oslo | 18:34 | |
| *** lbragstad is now known as lbragstad__ | 18:43 | |
| openstackgerrit | Ben Nemec proposed a change to openstack/oslo-incubator: Add lockutils fixture to OpportunisticTestCase https://review.openstack.org/79655 | 18:44 |
| openstackgerrit | Roman Podoliaka proposed a change to openstack/oslo-incubator: Add a base test case for DB schema comparison https://review.openstack.org/74081 | 18:47 |
| *** rpodolyaka has quit IRC | 18:51 | |
| *** rpodolyaka has joined #openstack-oslo | 18:52 | |
| *** harlowja has quit IRC | 19:04 | |
| *** sagaidakt has joined #openstack-oslo | 19:06 | |
| *** harlowja has joined #openstack-oslo | 19:08 | |
| *** sagaidakt has quit IRC | 19:09 | |
| *** lbragstad__ has quit IRC | 19:18 | |
| *** lbragstad has joined #openstack-oslo | 19:25 | |
| openstackgerrit | Zane Bitter proposed a change to openstack/oslo-incubator: Config generator: handle options in multiple groups https://review.openstack.org/79738 | 19:30 |
| *** saikrishna_ has quit IRC | 19:30 | |
| *** kozolup has joined #openstack-oslo | 19:31 | |
| *** beav has quit IRC | 19:33 | |
| *** beav has joined #openstack-oslo | 19:34 | |
| *** beav has quit IRC | 19:37 | |
| openstackgerrit | Doug Hellmann proposed a change to openstack/oslo.test: Fix up documentation files https://review.openstack.org/79742 | 19:40 |
| *** kozolup has quit IRC | 19:42 | |
| openstackgerrit | Doug Hellmann proposed a change to openstack/oslo.test: Fix up documentation files https://review.openstack.org/79742 | 19:47 |
| openstackgerrit | Doug Hellmann proposed a change to openstack/oslo.test: Fix up documentation files https://review.openstack.org/79742 | 19:53 |
| *** mriedem1 has joined #openstack-oslo | 20:02 | |
| *** mriedem has quit IRC | 20:03 | |
| *** rpodolyaka has quit IRC | 20:07 | |
| *** arnaud has quit IRC | 20:18 | |
| *** rpodolyaka has joined #openstack-oslo | 21:09 | |
| dhellmann | viktors_: ping me when you have a few minutes to talk about the plan for oslo.db | 21:17 |
| dhellmann | viktors_: please :-) | 21:17 |
| *** harlowja has quit IRC | 21:28 | |
| *** wingwj has quit IRC | 21:28 | |
| *** dims_ has quit IRC | 21:28 | |
| *** dshulyak has quit IRC | 21:28 | |
| *** jd__ has quit IRC | 21:28 | |
| *** zhiyan_ has quit IRC | 21:28 | |
| *** mriedem1 has quit IRC | 21:28 | |
| *** bknudson has quit IRC | 21:28 | |
| *** yamahata has quit IRC | 21:28 | |
| *** dhellmann has quit IRC | 21:28 | |
| *** clarkb has quit IRC | 21:28 | |
| *** rpodolyaka has quit IRC | 21:28 | |
| *** viktors_ has quit IRC | 21:28 | |
| *** yamahata_ has quit IRC | 21:28 | |
| *** flaper87 has quit IRC | 21:28 | |
| *** lbragstad has quit IRC | 21:28 | |
| *** jeblair has quit IRC | 21:28 | |
| *** openstackgerrit has quit IRC | 21:28 | |
| *** dteselkin has quit IRC | 21:28 | |
| *** SergeyLukjanov has quit IRC | 21:28 | |
| *** ruhe has quit IRC | 21:28 | |
| *** gilliard has quit IRC | 21:28 | |
| *** ihrachys has quit IRC | 21:28 | |
| *** sileht has quit IRC | 21:28 | |
| *** boris-42 has quit IRC | 21:28 | |
| *** russellb has quit IRC | 21:29 | |
| *** ekarlso has quit IRC | 21:29 | |
| *** bnemec has quit IRC | 21:29 | |
| *** anteaya has quit IRC | 21:29 | |
| *** hdd has quit IRC | 21:29 | |
| *** sileht has joined #openstack-oslo | 21:31 | |
| *** zhiyan_ has joined #openstack-oslo | 21:38 | |
| *** jd__ has joined #openstack-oslo | 21:38 | |
| *** jd__ is now known as 23LAAONZW | 21:41 | |
| *** russellb has joined #openstack-oslo | 21:41 | |
| *** dims_ has joined #openstack-oslo | 21:41 | |
| *** dshulyak has joined #openstack-oslo | 21:41 | |
| *** boris-42 has joined #openstack-oslo | 21:41 | |
| *** morganfainberg has joined #openstack-oslo | 21:41 | |
| *** mriedem has joined #openstack-oslo | 21:41 | |
| *** rpodolyaka has joined #openstack-oslo | 21:41 | |
| *** lbragstad has joined #openstack-oslo | 21:41 | |
| *** yamahata has joined #openstack-oslo | 21:41 | |
| *** viktors_ has joined #openstack-oslo | 21:41 | |
| *** bknudson has joined #openstack-oslo | 21:41 | |
| *** SergeyLukjanov has joined #openstack-oslo | 21:41 | |
| *** hdd has joined #openstack-oslo | 21:41 | |
| *** jeblair has joined #openstack-oslo | 21:41 | |
| *** clarkb has joined #openstack-oslo | 21:41 | |
| *** flaper87 has joined #openstack-oslo | 21:41 | |
| *** openstackgerrit has joined #openstack-oslo | 21:41 | |
| *** yamahata_ has joined #openstack-oslo | 21:41 | |
| *** dhellmann has joined #openstack-oslo | 21:41 | |
| *** ihrachys has joined #openstack-oslo | 21:41 | |
| *** dteselkin has joined #openstack-oslo | 21:41 | |
| *** ruhe has joined #openstack-oslo | 21:41 | |
| *** anteaya has joined #openstack-oslo | 21:41 | |
| *** ekarlso has joined #openstack-oslo | 21:41 | |
| *** gilliard has joined #openstack-oslo | 21:41 | |
| *** bnemec has joined #openstack-oslo | 21:41 | |
| *** 23LAAONZW is now known as jd__` | 21:41 | |
| *** jd__` is now known as jd__ | 21:42 | |
| *** mriedem has quit IRC | 21:44 | |
| *** harlowja has joined #openstack-oslo | 21:45 | |
| *** lbragstad has quit IRC | 21:47 | |
| openstackgerrit | Cyril Roelandt proposed a change to openstack/oslo-incubator: Python 3: enable more tests, remove a failing one https://review.openstack.org/79781 | 22:07 |
| openstackgerrit | Cyril Roelandt proposed a change to openstack/oslo-incubator: Python 3: enable more tests, remove a failing one https://review.openstack.org/79781 | 22:16 |
| *** flaper87 is now known as flaper87|afk | 22:19 | |
| *** dims_ has quit IRC | 22:30 | |
| bnemec | Crud, the requirements change failed a test in the gate for some probably unrelated reason. | 22:34 |
| *** yamahata has quit IRC | 22:43 | |
| *** openstackstatus has joined #openstack-oslo | 22:44 | |
| *** dims_ has joined #openstack-oslo | 22:45 | |
| dhellmann | bnemec: :-| | 22:54 |
| dhellmann | bnemec: it's past time to call it a night here, but if you're around when that finishes and can reverify it that would be great -- otherwise I'll do it first thing tomorrow | 22:55 |
| dhellmann | bnemec: dunno which tz you're in... | 22:55 |
| *** lbragstad has joined #openstack-oslo | 23:07 | |
| *** rpodolyaka has quit IRC | 23:09 | |
| *** rpodolyaka has joined #openstack-oslo | 23:12 | |
| *** mriedem has joined #openstack-oslo | 23:31 | |
| openstackgerrit | Zhongyue Luo proposed a change to openstack/oslo-incubator: Clean up test_imageutils https://review.openstack.org/78093 | 23:33 |
| openstackgerrit | Zhongyue Luo proposed a change to openstack/oslo-incubator: Remove redundant test cases in test_imageutils https://review.openstack.org/77030 | 23:33 |
| openstackgerrit | Zhongyue Luo proposed a change to openstack/oslo-incubator: Adds test condition in test_periodic https://review.openstack.org/72071 | 23:34 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!