| *** tsekiyama has quit IRC | 00:01 | |
| *** dims_ has quit IRC | 00:01 | |
| *** dims_ has joined #openstack-oslo | 00:01 | |
| *** tsekiyam_ has quit IRC | 00:03 | |
| *** dims_ has quit IRC | 00:06 | |
| openstackgerrit | lifeless proposed a change to openstack-dev/pbr: Teach pbr about post versioned dev versions. https://review.openstack.org/80449 | 00:16 |
|---|---|---|
| openstackgerrit | lifeless proposed a change to openstack-dev/pbr: Raise an error if preversion versions are too low https://review.openstack.org/108269 | 00:16 |
| openstackgerrit | lifeless proposed a change to openstack-dev/pbr: Look for and process sem-ver pseudo headers in git https://review.openstack.org/108270 | 00:17 |
| openstackgerrit | lifeless proposed a change to openstack-dev/pbr: Introduce a SemanticVersion object https://review.openstack.org/94108 | 00:17 |
| *** zzzeek has quit IRC | 00:22 | |
| *** arnaud has quit IRC | 00:23 | |
| *** dims_ has joined #openstack-oslo | 00:29 | |
| *** dims_ is now known as dims | 00:37 | |
| *** markmcclain has quit IRC | 00:50 | |
| *** Ish__ has joined #openstack-oslo | 00:52 | |
| *** arnaud has joined #openstack-oslo | 01:08 | |
| *** yamahata has joined #openstack-oslo | 01:10 | |
| openstackgerrit | A change was merged to openstack/oslo.vmware: Port the Datastore and DatastorePath objects https://review.openstack.org/105274 | 01:37 |
| *** Ish__ has quit IRC | 01:51 | |
| *** Krast has joined #openstack-oslo | 02:00 | |
| *** dims has quit IRC | 02:04 | |
| *** dims has joined #openstack-oslo | 02:05 | |
| *** dims is now known as Guest48695 | 02:05 | |
| *** Guest48695 has quit IRC | 02:09 | |
| *** Krast has quit IRC | 02:23 | |
| *** arnaud has quit IRC | 02:34 | |
| *** SridharG has joined #openstack-oslo | 02:34 | |
| *** mriedem has left #openstack-oslo | 02:36 | |
| *** mriedem has quit IRC | 02:36 | |
| *** arnaud has joined #openstack-oslo | 02:38 | |
| *** Krast has joined #openstack-oslo | 02:42 | |
| *** pcm_ has quit IRC | 02:49 | |
| openstackgerrit | YAMAMOTO Takashi proposed a change to openstack/oslo-incubator: Specify namedtuple_as_object=False when using simplejson https://review.openstack.org/113760 | 03:27 |
| openstackgerrit | Zhi Kun Liu proposed a change to openstack/oslo.messaging: Add parameter to customize Qpid receiver capacity https://review.openstack.org/112536 | 03:43 |
| *** praneshp has quit IRC | 03:55 | |
| openstackgerrit | A change was merged to openstack/oslo.messaging: Enable PEP8 check E241 https://review.openstack.org/109077 | 03:59 |
| *** arnaud has quit IRC | 04:02 | |
| *** jroll|dupe has joined #openstack-oslo | 04:03 | |
| *** arnaud has joined #openstack-oslo | 04:05 | |
| *** jroll has quit IRC | 04:05 | |
| *** jroll|dupe is now known as jroll | 04:05 | |
| *** arnaud has quit IRC | 04:08 | |
| *** arnaud has joined #openstack-oslo | 04:18 | |
| *** praneshp_ has joined #openstack-oslo | 04:32 | |
| *** ildikov has quit IRC | 04:36 | |
| *** jgrimm has quit IRC | 05:01 | |
| *** SridharG has quit IRC | 05:25 | |
| *** k4n0 has joined #openstack-oslo | 05:30 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/oslo.vmware: Imported Translations from Transifex https://review.openstack.org/113785 | 06:08 |
| *** HenryG_ has joined #openstack-oslo | 06:09 | |
| *** ildikov has joined #openstack-oslo | 06:11 | |
| *** HenryG has quit IRC | 06:12 | |
| *** SridharG has joined #openstack-oslo | 06:13 | |
| *** Alexei_987 has quit IRC | 06:14 | |
| *** stevemar has quit IRC | 06:42 | |
| *** stevemar has joined #openstack-oslo | 06:42 | |
| *** flaper87|afk is now known as flaper87 | 06:44 | |
| *** AAzza_afk is now known as AAzza | 06:47 | |
| *** stevemar has quit IRC | 06:47 | |
| *** arnaud has quit IRC | 06:48 | |
| *** arnaud has joined #openstack-oslo | 06:57 | |
| *** jaosorior has joined #openstack-oslo | 07:18 | |
| *** ihrachyshka has joined #openstack-oslo | 07:20 | |
| *** ihrachyshka has quit IRC | 07:23 | |
| *** Alexei_987 has joined #openstack-oslo | 07:27 | |
| *** noelbk has quit IRC | 07:51 | |
| openstackgerrit | Ma Wen Cheng proposed a change to openstack/oslo.messaging: Disable Topic name used as Queue name for QPID consumers https://review.openstack.org/113808 | 08:11 |
| *** arnaud has quit IRC | 08:12 | |
| *** wenchma has joined #openstack-oslo | 08:18 | |
| *** ihrachyshka has joined #openstack-oslo | 08:26 | |
| *** pblaho has joined #openstack-oslo | 09:08 | |
| *** pblaho has quit IRC | 09:11 | |
| *** pblaho has joined #openstack-oslo | 09:11 | |
| *** ihrachyshka has quit IRC | 09:11 | |
| d0ugal | Having some issue with my oslo.db migration. The options are not showing up in the generated config sample. What is the correct way to include them now? | 09:58 |
| *** sheeprine has quit IRC | 10:00 | |
| *** bnemec has quit IRC | 10:00 | |
| *** pblaho is now known as pblaho|afk | 10:01 | |
| *** ihrachyshka has joined #openstack-oslo | 10:07 | |
| *** sheeprine has joined #openstack-oslo | 10:07 | |
| *** bnemec has joined #openstack-oslo | 10:07 | |
| *** ihrachyshka has quit IRC | 10:07 | |
| *** ihrachyshka has joined #openstack-oslo | 10:07 | |
| *** pcm_ has joined #openstack-oslo | 10:13 | |
| *** pcm__ has joined #openstack-oslo | 10:15 | |
| *** pcm_ has quit IRC | 10:18 | |
| *** ihrachyshka has quit IRC | 10:23 | |
| *** sheeprine has quit IRC | 10:23 | |
| *** bnemec has quit IRC | 10:23 | |
| *** ihrachyshka has joined #openstack-oslo | 10:30 | |
| *** sheeprine has joined #openstack-oslo | 10:30 | |
| *** bnemec has joined #openstack-oslo | 10:30 | |
| *** ihrachyshka has quit IRC | 10:30 | |
| *** ihrachyshka has joined #openstack-oslo | 10:30 | |
| *** markmc has joined #openstack-oslo | 10:36 | |
| *** yamahata has quit IRC | 10:42 | |
| openstackgerrit | Rossella Sblendido proposed a change to openstack/oslo.db: Retry query if db deadlock error is received https://review.openstack.org/109549 | 10:44 |
| amrith | YorikSar, russellb , pblaho|afk , dhellmann ... would you please re-review https://review.openstack.org/#/c/110933/ | 10:48 |
| *** dims has joined #openstack-oslo | 11:03 | |
| *** dims is now known as Guest25021 | 11:03 | |
| *** pblaho|afk is now known as pblaho | 11:03 | |
| *** Guest25021 has quit IRC | 11:04 | |
| *** dims_ has joined #openstack-oslo | 11:05 | |
| *** dims_ has quit IRC | 11:22 | |
| *** dims_ has joined #openstack-oslo | 11:23 | |
| *** AAzza is now known as AAzza_afk | 11:24 | |
| *** dims__ has joined #openstack-oslo | 11:26 | |
| *** dims_ has quit IRC | 11:27 | |
| openstackgerrit | A change was merged to openstack/oslo.rootwrap: Refactor common parts from cmd to wrapper https://review.openstack.org/107083 | 11:30 |
| *** pblaho is now known as pblaho|afk | 11:31 | |
| *** pblaho|afk is now known as pblaho | 11:39 | |
| dhellmann | d0ugal: if you're using the config generator from the incubator, you probably need to add the -l option with oslo.db as an argument http://git.openstack.org/cgit/openstack/oslo-incubator/tree/tools/config/generate_sample.sh#n35 | 11:46 |
| d0ugal | dhellmann: aha! awesome, that is just what I've been looking for. | 11:48 |
| dhellmann | d0ugal: if you wanted to be hip, you could update to use the config generator in oslo.config, which uses that mode only and no longer parses code looking for option definitions | 11:53 |
| d0ugal | dhellmann: I did look at that, I'm not sure I'm quite hip enough :) | 11:54 |
| d0ugal | dhellmann: (maybe once I solve my other issues) | 11:54 |
| *** dims__ has quit IRC | 12:04 | |
| *** dims_ has joined #openstack-oslo | 12:05 | |
| *** dims_ has quit IRC | 12:09 | |
| openstackgerrit | Yuriy Taraday proposed a change to openstack/oslo.rootwrap: Add an option to run rootwrap as a daemon https://review.openstack.org/81798 | 12:21 |
| *** boris-42 has quit IRC | 12:21 | |
| *** dims_ has joined #openstack-oslo | 12:22 | |
| *** bknudson1 has quit IRC | 12:32 | |
| *** boris-42 has joined #openstack-oslo | 12:32 | |
| openstackgerrit | A change was merged to openstack-dev/cookiecutter: Use spaces instead of tabs for indentation https://review.openstack.org/103869 | 12:35 |
| *** dims_ has quit IRC | 12:39 | |
| *** dims_ has joined #openstack-oslo | 12:40 | |
| *** dims_ has quit IRC | 12:44 | |
| openstackgerrit | Rossella Sblendido proposed a change to openstack/oslo.db: Retry query if db deadlock error is received https://review.openstack.org/109549 | 12:45 |
| dhellmann | d0ugal: one step at a time :-) | 12:46 |
| openstackgerrit | Ihar Hrachyshka proposed a change to openstack/oslo-incubator: Add mock cleanup function to each test case https://review.openstack.org/95465 | 12:47 |
| *** harlowja_at_home has joined #openstack-oslo | 12:48 | |
| ihrachyshka | dhellmann: quite a lot of cores gave their +2's for mysqlconnector spec. can we proceed with pushing it? | 12:51 |
| *** bknudson has joined #openstack-oslo | 12:52 | |
| dhellmann | ihrachyshka: yep, I think the last time I looked they hadn't yet | 12:55 |
| dhellmann | ihrachyshka: regarding rpodolyaka_afk's point on the use of "switch", how about if we rename the blueprint after we approve the spec? maybe enable-mysql-connector or enable-alternate-db-drivers? | 12:57 |
| dhellmann | ihrachyshka: I can do all of that, once you and I agree on a new name, so just let me know | 12:57 |
| ihrachyshka | dhellmann: I'm for it both hands if it does trigger another month of collecting approvals :) | 12:57 |
| dhellmann | ihrachyshka: yeah, I'll just push the name change through on my own as PTL, so don't worry about that | 12:58 |
| dhellmann | ihrachyshka: enable-alternate-db-drivers? | 12:58 |
| dhellmann | well, maybe we should go with enable-mysql-connector for now to keep the scope in mind | 12:58 |
| * dhellmann needs to finish his tea before naming things | 12:59 | |
| ihrachyshka | dhellmann: ok, thanks :) well, enable-alternate-db-drivers is too wide I think. though I'll do the work so that it's reusable for other drivers, I'll test it with mysqlconnector before it's done. After that, we may try to check whether other drivers are easy to achieve. | 12:59 |
| ihrachyshka | +100 for enable-mysql-connector | 12:59 |
| dhellmann | ok, enable-mysql-connector it is then :-) | 12:59 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Use index-filter instead of tree-filter in graduate.sh https://review.openstack.org/109685 | 13:02 |
| openstackgerrit | A change was merged to openstack/oslo-specs: Switch from MySQLdb to MySQL Connector https://review.openstack.org/108355 | 13:02 |
| openstackgerrit | Doug Hellmann proposed a change to openstack/oslo-specs: Rename switch-to-mysql-connector spec https://review.openstack.org/113872 | 13:05 |
| openstackgerrit | A change was merged to openstack/oslo-specs: Rename switch-to-mysql-connector spec https://review.openstack.org/113872 | 13:08 |
| dhellmann | ihrachyshka: spec and blueprint approved and targeted | 13:09 |
| dhellmann | ihrachyshka: please keep the status up to date as you start work :-) | 13:09 |
| *** gordc has joined #openstack-oslo | 13:10 | |
| *** dims_ has joined #openstack-oslo | 13:10 | |
| openstackgerrit | A change was merged to openstack-dev/pbr: Introduce a SemanticVersion object https://review.openstack.org/94108 | 13:16 |
| ihrachyshka | dhellmann: what do you mean - keep the status up to date? | 13:16 |
| ihrachyshka | dhellmann: like 'report publicly the progress'? | 13:16 |
| dhellmann | ihrachyshka: no, like set the status field of the blueprint in launchpad if it's not being updated automatically | 13:17 |
| dhellmann | ihrachyshka: the "Implementation" field | 13:17 |
| ihrachyshka | ah, ok | 13:19 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Enhance versionutils.deprecated to work with classes https://review.openstack.org/104916 | 13:20 |
| *** jecarey has joined #openstack-oslo | 13:21 | |
| openstackgerrit | A change was merged to openstack-dev/oslo-cookiecutter: Use oslotest by default https://review.openstack.org/106876 | 13:24 |
| openstackgerrit | A change was merged to openstack-dev/oslo-cookiecutter: Run py33 tests before py2x https://review.openstack.org/106875 | 13:24 |
| *** stevemar has joined #openstack-oslo | 13:24 | |
| *** mriedem has joined #openstack-oslo | 13:29 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Correct coercion of logged message to unicode https://review.openstack.org/110772 | 13:32 |
| *** zzzeek has joined #openstack-oslo | 13:34 | |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Add unicode coercion of logged messages to ContextFormatter https://review.openstack.org/112135 | 13:34 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: log: make tests portable https://review.openstack.org/110297 | 13:34 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Except socket.error if syslog isn't running https://review.openstack.org/110753 | 13:36 |
| openstackgerrit | A change was merged to openstack/oslo-incubator: Set keystonemiddleware and routes.middleware to log on WARN level https://review.openstack.org/109379 | 13:37 |
| *** tongli has joined #openstack-oslo | 13:38 | |
| *** jecarey has quit IRC | 13:49 | |
| *** jecarey has joined #openstack-oslo | 13:50 | |
| openstackgerrit | A change was merged to openstack-dev/pbr: add docs env to tox https://review.openstack.org/113529 | 13:51 |
| openstackgerrit | gordon chung proposed a change to openstack/oslo-specs: graduate-oslo-middleware https://review.openstack.org/110353 | 13:52 |
| *** harlowja_at_home has quit IRC | 13:54 | |
| *** zzzeek has quit IRC | 13:58 | |
| openstackgerrit | A change was merged to openstack-dev/pbr: cleanup tox.ini https://review.openstack.org/113553 | 14:11 |
| openstackgerrit | A change was merged to openstack/oslo.messaging: Enable PEP8 check E265 https://review.openstack.org/109079 | 14:11 |
| openstackgerrit | A change was merged to openstack/oslo.messaging: Enable PEP8 check E714 https://review.openstack.org/109080 | 14:11 |
| *** Krast has quit IRC | 14:13 | |
| *** Krast has joined #openstack-oslo | 14:14 | |
| *** AAzza_afk is now known as AAzza | 14:16 | |
| *** openstackgerrit has quit IRC | 14:16 | |
| *** mriedem has quit IRC | 14:17 | |
| *** openstackgerrit has joined #openstack-oslo | 14:17 | |
| *** jecarey has quit IRC | 14:18 | |
| *** mriedem has joined #openstack-oslo | 14:26 | |
| *** Krast has quit IRC | 14:28 | |
| *** Krast has joined #openstack-oslo | 14:28 | |
| openstackgerrit | A change was merged to openstack/oslo.rootwrap: Move test requirement coverage into tox.ini https://review.openstack.org/108345 | 14:29 |
| *** Krast has quit IRC | 14:31 | |
| *** Krast has joined #openstack-oslo | 14:31 | |
| *** markmcclain has joined #openstack-oslo | 14:48 | |
| *** tsekiyama has joined #openstack-oslo | 14:49 | |
| *** markmcclain1 has joined #openstack-oslo | 14:51 | |
| *** flaper87 is now known as flaper87|afk | 14:54 | |
| *** markmcclain has quit IRC | 14:55 | |
| *** linkid has joined #openstack-oslo | 14:57 | |
| *** pblaho is now known as pblaho|afk | 14:58 | |
| linkid | hi | 14:58 |
| linkid | I have a question about oslo.messaging | 14:58 |
| YorikSar | amrith: will do | 14:59 |
| linkid | is it possible to use the rpc communication without a bloquing server ? | 14:59 |
| *** jgrimm has joined #openstack-oslo | 15:00 | |
| linkid | 'cause I used an rpc server with a callback, but I had to use server.wait() to reveive the client message, which is a blocked method | 15:02 |
| *** ildikov has quit IRC | 15:06 | |
| ihrachyshka | linkid: which executor do you use? | 15:09 |
| linkid | eventlet | 15:09 |
| *** jecarey has joined #openstack-oslo | 15:09 | |
| *** pblaho|afk is now known as pblaho | 15:10 | |
| *** yamahata has joined #openstack-oslo | 15:19 | |
| *** noelbk has joined #openstack-oslo | 15:32 | |
| *** jaosorior has quit IRC | 15:32 | |
| openstackgerrit | Christian Berendt proposed a change to openstack/oslo.rootwrap: Replace string format arguments with function parameters https://review.openstack.org/94378 | 15:38 |
| *** SridharG has quit IRC | 15:42 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/oslo-incubator: add list_opts to all modules with configuration options https://review.openstack.org/113940 | 15:44 |
| *** pcm__ has quit IRC | 15:57 | |
| *** ihrachyshka has quit IRC | 15:58 | |
| krotscheck | I have a question on oslo.config - does it matter in what order CONF(project) and CONF.register_opts(FOO, ‘foocategory’) are called? | 15:59 |
| *** k4n0 has quit IRC | 16:01 | |
| *** AAzza is now known as AAzza_afk | 16:03 | |
| *** noelbk has quit IRC | 16:09 | |
| *** praneshp has joined #openstack-oslo | 16:09 | |
| *** praneshp_ has quit IRC | 16:09 | |
| markmc | krotscheck, nope | 16:12 |
| krotscheck | markmc: Thanks | 16:12 |
| markmc | krotscheck, you have to call register_cli_opts() before CONF(), but not register_opts() | 16:12 |
| krotscheck | Awesome, thanks | 16:12 |
| markmc | np | 16:12 |
| YorikSar | ttx: Hi, have a minute? | 16:14 |
| YorikSar | ttx: I'm trying to figure out a way to test proper cleanup on process termination for my oslo.rootwrap patch and can't think of any. | 16:15 |
| YorikSar | The problem was that _shutdown never run in atexit handler because I didn't specify exitpriority to Finalizer class (from multiprocessing). | 16:16 |
| YorikSar | To test that I should run a separate process, let it spawn another (rootwrap) process, and then die. And then I should somehow see if my grandchild is alive or dead. | 16:17 |
| YorikSar | I don't think there's a way to do it w/o some psutils or smth. | 16:17 |
| YorikSar | ttx: Oh, you're not online. | 16:18 |
| YorikSar | ajo: ^ may be you can think of smth? | 16:19 |
| ajo | hi YorikSar , 1 sec, reading | 16:19 |
| *** arnaud has joined #openstack-oslo | 16:21 | |
| ajo | YorikSar, what do you mean by "proper cleanup" ? :) | 16:21 |
| ajo | ah, the atexit stuff | 16:22 |
| ajo | let me look at your last changes | 16:22 |
| YorikSar | ajo: ttx noticed that benchmark run leaves a rootwrap process behind. It should be shut down on client process exit (finalizers are run in atexit handler). | 16:22 |
| YorikSar | ajo: And that didn't happend because I missed one option to multiprocessing version of finalizers. | 16:23 |
| ajo | aha | 16:23 |
| ajo | so for actually checking that, you need to spawn a process, that spawns a rootwrap | 16:23 |
| ajo | that spawns a process | 16:23 |
| ajo | yikes | 16:23 |
| *** praneshp has quit IRC | 16:23 | |
| ajo | and then check that when first process exits, the rootwrap is gone | 16:24 |
| YorikSar | Yep | 16:24 |
| YorikSar | From what I understand, that's rather hard... | 16:24 |
| ajo | don't you have an small implementation example to start and stop? | 16:25 |
| ajo | you may need to list processes, yes | 16:25 |
| ajo | rootwrap testing is unit, or is it functional? | 16:25 |
| ajo | it sounds like a functional test | 16:25 |
| YorikSar | There are 2 tests that verify that rootwrap process exits, but they rely on wait() system call that waits on children only. | 16:26 |
| *** yamahata has quit IRC | 16:26 | |
| YorikSar | Daemon-related code is covered with functional tests as unit tests won't make a lot of sense with it. | 16:26 |
| YorikSar | (btw, I've added some functional testing for usual rootwrap as well recently) | 16:27 |
| *** Ish__ has joined #openstack-oslo | 16:29 | |
| openstackgerrit | Yuriy Taraday proposed a change to openstack/oslo.rootwrap: Add an option to run rootwrap as a daemon https://review.openstack.org/81798 | 16:43 |
| *** AAzza_afk is now known as AAzza | 16:47 | |
| *** praneshp has joined #openstack-oslo | 16:53 | |
| openstackgerrit | Yuriy Taraday proposed a change to openstack/oslo.rootwrap: Add daemon mode to benchmark https://review.openstack.org/107085 | 16:53 |
| *** pcm_ has joined #openstack-oslo | 16:56 | |
| *** alexpilotti has joined #openstack-oslo | 16:57 | |
| *** AAzza is now known as AAzza_afk | 17:05 | |
| *** YorikSar has quit IRC | 17:07 | |
| *** YorikSar has joined #openstack-oslo | 17:09 | |
| *** zzzeek has joined #openstack-oslo | 17:11 | |
| *** yamahata has joined #openstack-oslo | 17:11 | |
| *** nealph has joined #openstack-oslo | 17:15 | |
| *** praneshp has quit IRC | 17:15 | |
| *** Alexei_987 has quit IRC | 17:17 | |
| *** yamahata has quit IRC | 17:18 | |
| openstackgerrit | YAMAMOTO Takashi proposed a change to openstack/oslo.serialization: Specify namedtuple_as_object=False when using simplejson https://review.openstack.org/113973 | 17:19 |
| *** praneshp has joined #openstack-oslo | 17:22 | |
| *** ajo has quit IRC | 17:23 | |
| nealph | question on oslo messaging: I'm digging for any capability to throttle message frequency (i.e. to avoid input flooding to downstream subscribers...in this case Ceilometer). Is there any such capability someone can point me to? | 17:23 |
| *** markmc has quit IRC | 17:27 | |
| dhellmann | nealph: I don't know if we have that in the messaging library. You might be able to do it through the broker, though? | 17:30 |
| openstackgerrit | Dmitry Tantsur proposed a change to openstack/oslo-incubator: Add field_labels parameter to cliutils.print_list https://review.openstack.org/113978 | 17:40 |
| nealph | dhellman: yep, thinking on that approach too. I see that qpid and rabbitmq both implement resource-based flow control but was curious to know if anyone had worked through the logic to do that in the "producers", i.e. oslo messaging for most services. | 17:43 |
| nealph | I'll keep nosing around. | 17:44 |
| *** celttechie has joined #openstack-oslo | 17:44 | |
| dhellmann | nealph: with notifications I think we wouldn't want the producer blocked, so you'd want to do the throttling after the message ends up safely in a queue somewhere | 17:46 |
| *** linkid has left #openstack-oslo | 17:48 | |
| openstackgerrit | YAMAMOTO Takashi proposed a change to openstack/oslo.serialization: Specify namedtuple_as_object=False when using simplejson https://review.openstack.org/113973 | 17:51 |
| *** Ish__ has quit IRC | 17:55 | |
| *** arnaud has quit IRC | 17:59 | |
| *** pblaho has quit IRC | 18:09 | |
| *** AAzza_afk is now known as AAzza | 18:16 | |
| mriedem | gordc: i don't believe 'reverify' is a thing | 18:17 |
| mriedem | gordc: like 'reverify no bug' isn't a thing | 18:18 |
| *** dims_ has quit IRC | 18:19 | |
| *** dims_ has joined #openstack-oslo | 18:19 | |
| openstackgerrit | Ihar Hrachyshka proposed a change to openstack/oslo.db: Set utf8 encoding for mysql and postgresql https://review.openstack.org/111236 | 18:20 |
| *** dims_ has quit IRC | 18:21 | |
| *** Alexei_987 has joined #openstack-oslo | 18:21 | |
| *** dims has joined #openstack-oslo | 18:21 | |
| gordc | mriedem: hmm... i've seen people retriggering jobs like that. i was just too lazy to search out the real bug and assume there was a bug in ci which allowed you to recheck without bug # | 18:23 |
| mriedem | gordc: you can now recheck w/o a bug number | 18:24 |
| mriedem | gordc: but i don't think that extends to reverify | 18:24 |
| *** krotscheck has quit IRC | 18:24 | |
| gordc | mriedem: yeah that's what i thought. i feel like there's a bug that got in that makes it work for reverify too... or it seems to restart tests when i've done it. | 18:25 |
| *** krotscheck has joined #openstack-oslo | 18:25 | |
| gordc | mriedem: or do you mean it's not really reverifying when i type reverify...ie. it's just reposting same results? | 18:25 |
| *** arnaud has joined #openstack-oslo | 18:26 | |
| mriedem | gordc: oh i guess it is running tests again | 18:27 |
| mriedem | hm, | 18:27 |
| mriedem | gordc: well it'd be nice to see what's causing the failures rather than reverifying something through the gate | 18:27 |
| mriedem | maybe the code is racing | 18:27 |
| mriedem | gordc: http://logs.openstack.org/99/99599/3/check/check-grenade-dsvm/45be8d5/logs/grenade.sh.txt.gz#_2014-08-13_14_11_44_105 | 18:28 |
| gordc | mriedem: yeah... i figured i'd try one more time before i find time to actually dig into it | 18:28 |
| mriedem | http://logs.openstack.org/99/99599/3/check/check-grenade-dsvm/45be8d5/logs/new/screen-ceilometer-api.txt.gz | 18:29 |
| mriedem | yeah, i'd -2 that | 18:29 |
| *** AAzza is now known as AAzza_afk | 18:29 | |
| gordc | mriedem: interesting... didn't see that | 18:30 |
| gordc | mriedem: i rebased it...i don't feel like -2ing myself. :) | 18:30 |
| gordc | mriedem: looks like i got to think on this rebase. :( thanks for pointing it out. | 18:31 |
| mriedem | gordc: np | 18:31 |
| *** dims_ has joined #openstack-oslo | 18:32 | |
| *** praneshp_ has joined #openstack-oslo | 18:33 | |
| *** pcm__ has joined #openstack-oslo | 18:34 | |
| zzzeek | der, review.openstack borked ? | 18:36 |
| zzzeek | wow how strange. | 18:36 |
| *** jokke__ has joined #openstack-oslo | 18:36 | |
| gordc | zzzeek: back for me... with some ui tweaks apparently. | 18:36 |
| zzzeek | restart i guess | 18:37 |
| *** arnaud has quit IRC | 18:41 | |
| *** praneshp has quit IRC | 18:41 | |
| *** dims has quit IRC | 18:41 | |
| *** pcm_ has quit IRC | 18:41 | |
| *** openstackgerrit has quit IRC | 18:41 | |
| *** jokke_ has quit IRC | 18:41 | |
| *** praneshp_ is now known as praneshp | 18:41 | |
| *** AAzza_afk is now known as AAzza | 18:44 | |
| amrith | YorikSar, thx for comments, will push a patch set later today. | 18:44 |
| *** openstackgerrit has joined #openstack-oslo | 18:49 | |
| *** AAzza is now known as AAzza_afk | 18:51 | |
| *** krotscheck has quit IRC | 18:54 | |
| *** krotscheck has joined #openstack-oslo | 18:55 | |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Repair pysqlite transaction support https://review.openstack.org/113152 | 18:56 |
| openstackgerrit | Michael Bayer proposed a change to openstack/oslo.db: Repair pysqlite transaction support https://review.openstack.org/113152 | 18:57 |
| *** ildikov has joined #openstack-oslo | 19:03 | |
| *** ildikov has quit IRC | 19:08 | |
| *** ildikov has joined #openstack-oslo | 19:21 | |
| *** celttechie has quit IRC | 19:31 | |
| openstackgerrit | Christian Berendt proposed a change to openstack/oslo.db: Add doc8 to tox environment docs https://review.openstack.org/114003 | 19:46 |
| *** celttechie has joined #openstack-oslo | 19:51 | |
| openstackgerrit | Christian Berendt proposed a change to openstack/oslo.config: Add doc8 to tox environment docs https://review.openstack.org/114006 | 19:52 |
| *** zzzeek_ has joined #openstack-oslo | 20:06 | |
| *** zzzeek has quit IRC | 20:06 | |
| *** zzzeek_ is now known as zzzeek | 20:06 | |
| *** Ish__ has joined #openstack-oslo | 20:26 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/oslo-incubator: add list_opts to all modules with configuration options https://review.openstack.org/113940 | 20:27 |
| openstackgerrit | gordon chung proposed a change to openstack/oslo-specs: graduate-oslo-middleware https://review.openstack.org/110353 | 20:31 |
| openstackgerrit | gordon chung proposed a change to openstack/oslo.middleware: drop middleware suffix in class names https://review.openstack.org/114016 | 20:34 |
| *** Ish__ has quit IRC | 20:37 | |
| *** celttechie has quit IRC | 20:38 | |
| *** stevemar has quit IRC | 20:47 | |
| *** Ish__ has joined #openstack-oslo | 20:54 | |
| *** Ish__ has left #openstack-oslo | 20:55 | |
| openstackgerrit | gordon chung proposed a change to openstack/oslo.middleware: drop middleware suffix in class names https://review.openstack.org/114016 | 21:04 |
| *** celttechie has joined #openstack-oslo | 21:07 | |
| *** Alexei_987 has quit IRC | 21:12 | |
| *** wenchma has quit IRC | 21:12 | |
| *** YorikSar has quit IRC | 21:13 | |
| *** wenchma has joined #openstack-oslo | 21:13 | |
| *** YorikSar has joined #openstack-oslo | 21:14 | |
| *** Alexei_987 has joined #openstack-oslo | 21:15 | |
| *** arnaud__ has joined #openstack-oslo | 21:20 | |
| openstackgerrit | Brant Knudson proposed a change to openstack/oslo.messaging: Set sample_default for rpc_zmq_host https://review.openstack.org/114031 | 21:29 |
| *** jecarey has quit IRC | 21:40 | |
| *** HenryG_ is now known as HenryG | 21:57 | |
| *** mriedem has left #openstack-oslo | 21:58 | |
| *** jgrimm has quit IRC | 22:44 | |
| *** bknudson has quit IRC | 22:45 | |
| openstackgerrit | A change was merged to openstack/oslo.serialization: Specify namedtuple_as_object=False when using simplejson https://review.openstack.org/113973 | 22:48 |
| morganfainberg | dhellmann, zzzeek, if i need to change a column type (and transform/reinsert the data), is it safest (low volume table) to just re-create the table and port the data over? or is there some preferred "migration" magic that seems to work with everything nice and cleanly | 22:51 |
| morganfainberg | dhellmann, zzzeek, unfortunately this is migrate not alembic | 22:51 |
| zzzeek | well make a new column on the same table | 22:51 |
| zzzeek | then do an UPDATE | 22:51 |
| zzzeek | then drop the old column | 22:51 |
| morganfainberg | i thought sqlite didn't like renaming columns | 22:51 |
| * morganfainberg can't ever remeber what sqlite sucks at and doesn't suck at | 22:52 | |
| zzzeek | morganfainberg: ah SQLite. then migrate has this whole drop and recerate thing it should do automaticallyt | 22:52 |
| *** markmcclain1 has quit IRC | 22:52 | |
| zzzeek | migrate currently has better support for that :) | 22:52 |
| morganfainberg | yeah it needs to work on mysql, postgres, sqlite, and db2 | 22:52 |
| morganfainberg | i think those are all the dbs that officially work with keystone | 22:53 |
| zzzeek | in theory it should do the right thing | 22:53 |
| morganfainberg | ok so just add the column, port the data over, drop old column and rename? | 22:53 |
| zzzeek | if you have an explicit transformation though, then you’d need to have the two columns | 22:53 |
| openstackgerrit | YAMAMOTO Takashi proposed a change to openstack/oslo-incubator: Specify namedtuple_as_object=False when using simplejson https://review.openstack.org/113760 | 22:53 |
| zzzeek | morganfainberg: it “depends” | 22:53 |
| morganfainberg | yeah need to go from DATETIME to VARCHAR (with isotime str) | 22:53 |
| zzzeek | morganfainberg: on sqlite, table has to be rebuilt entirely, yes. migrate does this (somehow) | 22:53 |
| zzzeek | morganfainberg: other databases, dont know that migrate does that. you create a new col, emit UPDATE, drop the old | 22:54 |
| morganfainberg | because mysql is crappy about DATETIME(6) | 22:54 |
| zzzeek | morganfainberg: um also …what? whats wrong with DATETIME ? | 22:54 |
| morganfainberg | datetime in older than 5.6.4 of mysql doesn't have microsecond resolution | 22:54 |
| zzzeek | how about TIMESTAMP ? | 22:55 |
| morganfainberg | keystone relies on microsecond resolution for a couple things and it's stored in the DB | 22:55 |
| morganfainberg | same issue i think. | 22:55 |
| zzzeek | so you’d want to store an ISO string and not, say, epoch in millis ? | 22:55 |
| zzzeek | morganfainberg: also, SQLite doesnt have a DATETIME type. SQLAlchemy is already storing an ISO string in it | 22:55 |
| morganfainberg | looks like 5.6.4 adds subsecond resolution to TIME DATETIME and TIMESTAMP | 22:56 |
| morganfainberg | so, blah. need to go varchar. | 22:56 |
| morganfainberg | zzzeek, right but i actually need to declare this as varchar now and do the transform explicitly . i would rather not have wildly differing schemas on different RDBMSs (e.g.varchar and magic to convert for mysql only) | 22:57 |
| morganfainberg | though... i *guess* i could do that | 22:57 |
| zzzeek | morganfainberg: OK I htink you should have the same type on all backends | 22:58 |
| zzzeek | morganfainberg: but on SQLite, there is no DATETIME, it is already an ISO string | 22:58 |
| morganfainberg | zzzeek, right, but SQLAlchemy is doing the magic for me, right? | 22:59 |
| zzzeek | morganfainberg: but, if you stored as epoch in millis, then that could be everywhere. depends | 22:59 |
| zzzeek | morganfainberg: with the SQLite DATETIME, yes. | 22:59 |
| morganfainberg | yeah. | 22:59 |
| zzzeek | morganfainberg: with any system, you can make your own type that does the conversion | 22:59 |
| morganfainberg | that was what i was going to do | 22:59 |
| zzzeek | morganfainberg: so you could have it be mysql-only | 22:59 |
| morganfainberg | but just use it globally. | 22:59 |
| morganfainberg | rather than be "mysql" specific | 22:59 |
| morganfainberg | isotime_str_backed_datetime which would do timeutils.isotime() on store and timeutils.parse_isotime() on load | 23:00 |
| morganfainberg | for example | 23:00 |
| zzzeek | you can have a type that does somehing different on mysql, if you wanted | 23:01 |
| morganfainberg | zzzeek, would it be betteroverall to just be mysql specific in that case? | 23:01 |
| morganfainberg | zzzeek, i'd rather go with best practices vs. "eh something that works" | 23:02 |
| morganfainberg | :) | 23:02 |
| zzzeek | morganfainberg: with this issue i’m mostly thinking of, why is keystone comparing on time in millis and not a monotonically incrmenting counter of some kind :) | 23:02 |
| zzzeek | morganfainberg: timestamps for distinguishing events is a bad idea :) | 23:02 |
| morganfainberg | zzzeek, token expiration, tokens expire at isotime.subsectond | 23:02 |
| *** krotscheck has quit IRC | 23:03 | |
| morganfainberg | revocation events can invalidate tokens based on their expiration time | 23:03 |
| morganfainberg | 1 second resolution causes transient bugs | 23:03 |
| zzzeek | morganfainberg: that’s the funny part | 23:03 |
| *** krotscheck has joined #openstack-oslo | 23:03 | |
| morganfainberg | it's because many many more tokens match at 1s resolution | 23:03 |
| zzzeek | morganfainberg: when a too-great second resolution causes bugs that means something is comparing that timestamp to some other timestamp | 23:03 |
| morganfainberg | it is generally deemed that subsecond resolution (6 digits) is a narrow enough scope to prevent minimal transient bugs | 23:04 |
| zzzeek | morganfainberg: smells off to me, but i dont know the details | 23:05 |
| morganfainberg | so we issue 3 tokens that expire at 2014-08-04 00:00:00 with differing subsecond transactions. we issue a revocation event for one of them. but mysql trims it down to the full second | 23:05 |
| morganfainberg | all three tokens are now invalid | 23:05 |
| zzzeek | morganfainberg: i just find datetimes as strings really distasteful | 23:05 |
| morganfainberg | if mysql didn't trim the subseconds off only one token would be invalidated | 23:05 |
| zzzeek | ok wait | 23:06 |
| zzzeek | “issue a revocation event for one of them" | 23:06 |
| morganfainberg | we are trying to compare datetime to datetime, but the RDBMS doesn't keep subsecond resolution | 23:06 |
| zzzeek | and “one of them” means, “the one with timestamp X” ? that is, you are using timestamp as primary key | 23:06 |
| morganfainberg | with expiration time of X | 23:06 |
| zzzeek | see, thats the design failure | 23:06 |
| morganfainberg | we can't use token ids | 23:06 |
| morganfainberg | the token ids (the PK) are security and could be abused between event and when token is revoked at an endpoint | 23:07 |
| zzzeek | morganfainberg: would it be correct to have the subsecond resolution and then make this column a unique constraint ? | 23:07 |
| zzzeek | morganfainberg: if token ids are secure then there’s a one-way encrypted version that’s exposed.... | 23:07 |
| morganfainberg | we technically can't make it unique... but that is the design. | 23:07 |
| morganfainberg | zzzeek, the hash and the full id can be used interchanably | 23:08 |
| morganfainberg | it's a real headache | 23:08 |
| *** markmcclain has joined #openstack-oslo | 23:08 | |
| morganfainberg | i've been advocating using a non-auth granting id that can be used | 23:08 |
| morganfainberg | but i was shot down on that front | 23:08 |
| morganfainberg | we also use expiration_time to identify a chain of tokens, X token is rescoped which means Y token is issued. both contain the same expiration time | 23:08 |
| * morganfainberg would *rather* not use expiration_time in this case | 23:09 | |
| zzzeek | morganfainberg: why cant this just be a uuid4() value ? is that what your proposal was ? | 23:09 |
| morganfainberg | zzzeek, yep. but it introduces a backwards incompatible change | 23:10 |
| zzzeek | seems like this is all already a done deal. but the design is wrong :) | 23:10 |
| morganfainberg | zzzeek, among "it adds more data to the token" and "why do we need to do that" | 23:10 |
| zzzeek | “its a unique key but we techincally cant make it unique” :) run run run ! | 23:10 |
| zzzeek | tahts like php programming 101 :) | 23:10 |
| morganfainberg | zzzeek, yeah the reason we can't make it unique is there is a *small* chance with multiple keystone running that expiration times can collide | 23:11 |
| morganfainberg | i made this argument before and lost | 23:11 |
| zzzeek | but if you used GUIDs, then it wouldnt | 23:11 |
| zzzeek | too bad | 23:11 |
| morganfainberg | since microsecond resolution is only valid on a single clock domain within a system, due to skew | 23:11 |
| openstackgerrit | Clark Boylan proposed a change to openstack/oslo-incubator: Set python hash seed to 0 in tox.ini https://review.openstack.org/114052 | 23:11 |
| morganfainberg | yeah. it makes me kinda sad. | 23:11 |
| morganfainberg | i might try just flying the uuid thing again and wait for a -2 | 23:12 |
| zzzeek | well ill +1 it if you do | 23:12 |
| morganfainberg | because i really think expiration time is kinda broken. | 23:12 |
| zzzeek | til then i guess use the string | 23:12 |
| morganfainberg | zzzeek, and don't get me started about hashing and insecure hashing concerns | 23:13 |
| morganfainberg | zzzeek, e.g. you can *never use md5 ever* even if it isn't used for secure data because things could collide and auditors don't care if you aren't doing anything special with it, you're using it. same hold true for anything less than sha256 | 23:14 |
| zzzeek | yeah i generally stick to sha | 23:14 |
| zzzeek | but really, jsut uuid | 23:14 |
| zzzeek | good enough | 23:14 |
| morganfainberg | i tried that, and got told sha1 wasn't sufficient anymore unless it's sha256 | 23:14 |
| morganfainberg | or greater | 23:14 |
| morganfainberg | *great* | 23:15 |
| zzzeek | you can use uuid5() , it is based on a seed | 23:15 |
| zzzeek | its deterministic | 23:15 |
| zzzeek | passes all the fancy tests | 23:15 |
| morganfainberg | nope, you can't, it's sha1 based | 23:15 |
| zzzeek | oh it is? silly :) | 23:15 |
| morganfainberg | ep | 23:15 |
| morganfainberg | yep | 23:15 |
| zzzeek | i mean, not for security | 23:15 |
| zzzeek | but for collisions? | 23:15 |
| zzzeek | they think a uuid is going to collide ? | 23:15 |
| morganfainberg | uuid5 they think will and auditors throw a fit | 23:15 |
| morganfainberg | uuid4 is safe, it's random | 23:16 |
| * morganfainberg facepalms in the corner | 23:16 | |
| zzzeek | ok but you understand, that instead, you are using a value that *definitely* collides *all the time* | 23:16 |
| morganfainberg | i tried to pitch using uuid5 for user_ids in keystone, got told "nope we need the data to be hashed to sha256" | 23:16 |
| zzzeek | a timestamp! | 23:16 |
| morganfainberg | in this case yes. | 23:16 |
| zzzeek | whoever is hitting you with the sha256 show them this | 23:17 |
| zzzeek | and then tell them they are idiots | 23:17 |
| zzzeek | that they are nazis about sha256 but are totlaly fine with timestamps as not-quite-unique keys | 23:17 |
| zzzeek | sorry, it is just so stupid | 23:17 |
| morganfainberg | this whole problem stems from some technical debt that is a lot worse than i thought it was going to be in icehouse (not my code, but... still i let it pass into keystone) | 23:17 |
| morganfainberg | zzzeek, i've been working to go around the issues that make auditors unhappy (especially in keystone) most of this cycle i think. | 23:18 |
| zzzeek | its one thing for people to be overly paranoid, but for them to be completely inconsistent about it is insane | 23:18 |
| morganfainberg | they are consistent, sha1 is nogo | 23:18 |
| morganfainberg | ever at all in the application | 23:18 |
| morganfainberg | :P | 23:18 |
| zzzeek | its not consistent b.c. the rationale is one that they are willing to crap all over in other equivalent cases | 23:18 |
| zzzeek | id push to get these auditors to be aware that timestmaps collide | 23:19 |
| morganfainberg | it's hashing. if i *hashed* the timestamps and stored that in the db and only used it for the revocation it would be shot down | 23:19 |
| zzzeek | put a big comment, “dear auditors, these collide *all the time*!!” | 23:19 |
| morganfainberg | i think i'm just going to inject uuid4 into the token | 23:19 |
| morganfainberg | and revoke on that. | 23:19 |
| morganfainberg | if that doesn't exist fall back to the timestamp brokenness *backwards compat* | 23:19 |
| morganfainberg | and in ... 15kajillion releases maybe the brokenness can be removed | 23:20 |
| * morganfainberg is feeling a little bitter :P | 23:20 | |
| morganfainberg | zzzeek, thanks for your help dude, i appreciate it. | 23:20 |
| zzzeek | good luck | 23:20 |
| morganfainberg | zzzeek, you gave me the information i needed if i need to muck with the db schema | 23:20 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/oslo-incubator: Updated from global requirements https://review.openstack.org/114063 | 23:23 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/oslo.middleware: Updated from global requirements https://review.openstack.org/114064 | 23:23 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/oslo.serialization: Updated from global requirements https://review.openstack.org/114065 | 23:24 |
| openstackgerrit | Clark Boylan proposed a change to openstack/oslo.messaging: Make tests pass with random python hashseed. https://review.openstack.org/114079 | 23:26 |
| *** clarkb has joined #openstack-oslo | 23:36 | |
| *** gordc has quit IRC | 23:36 | |
| clarkb | hello oslo | 23:36 |
| clarkb | as part of my prep to upgrade tox I tested tox 1.7.2 against oslo* and https://review.openstack.org/#/c/114079/1 and https://review.openstack.org/#/c/114052/ come out of that | 23:37 |
| clarkb | if my change in 114079 does actually expose a real bug we should probably propose the equivalent of 114052 aganst messaging and fix that bug separately | 23:37 |
| clarkb | -incubator's failures were a bit more complicated and the py33 tests were already setting hashseed=0 so I didn't bother to try and fix them (that said we really should be able to run with a random hashseed because python3 defaults to that and telling people they need to set a different hashseed to run opesntack is :( ) | 23:41 |
| krotscheck | I know this is only tangentially related to oslo, however does anyone know why I wouldn’t be able to connect to rabbitmq unless the user in question was an admin? | 23:49 |
| *** markmcclain1 has joined #openstack-oslo | 23:56 | |
| *** markmcclain has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!