Monday, 2015-11-30

« Saturday, 2015-11-28 Index Tuesday, 2015-12-01 »
*** openstack has joined #openstack-barbican14:13
*** openstack has quit IRC14:13
*** openstack has joined #openstack-barbican14:16
*** openstack has quit IRC14:16
*** openstack has joined #openstack-barbican14:21
*** openstack has quit IRC14:22
*** openstack has joined #openstack-barbican14:23
*** openstack has joined #openstack-barbican14:26
*** openstack has quit IRC14:26
*** openstack has joined #openstack-barbican14:35
*** Stanzi has joined #openstack-barbican14:37
*** openstack has joined #openstack-barbican15:43
*** darrenmoffat has joined #openstack-barbican15:47
*** Stanzi_ has joined #openstack-barbican15:48
*** stanzi has quit IRC15:52
*** stanzi has joined #openstack-barbican15:53
*** stanzi has quit IRC15:55
*** stanzi has joined #openstack-barbican15:56
*** Kevin_Zheng has quit IRC15:56
*** Kevin_Zheng has joined #openstack-barbican15:56
openstackgerritKaitlin Farr proposed openstack/barbican: Add reno for release notes management  https://review.openstack.org/24891816:00
*** stanzi has quit IRC16:06
*** Stanzi_ has quit IRC16:07
*** stanzi has joined #openstack-barbican16:07
*** Stanzi_ has joined #openstack-barbican16:07
openstackgerritFernando Diaz proposed openstack/castellan: Add created property to Managed Objects  https://review.openstack.org/23815016:09
*** lisaclark1 has quit IRC16:09
*** Stanzi_ has quit IRC16:15
*** stanzi has quit IRC16:15
*** openstackstatus has joined #openstack-barbican16:15
*** ChanServ sets mode: +v openstackstatus16:15
*** lisaclark1 has joined #openstack-barbican16:17
openstackgerritKaitlin Farr proposed openstack/barbican: Add reno for release notes management  https://review.openstack.org/24891816:20
*** rhagarty has joined #openstack-barbican16:27
*** Stanzi has joined #openstack-barbican16:30
*** stanzi_ has joined #openstack-barbican16:30
*** stanzi_ has quit IRC16:35
*** Stanzi has quit IRC16:35
*** chlong has joined #openstack-barbican16:40
*** Stanzi has joined #openstack-barbican16:46
*** Stanzi has quit IRC16:51
*** stanzi has joined #openstack-barbican17:01
*** Stanzi_ has joined #openstack-barbican17:02
*** jaosorior has quit IRC17:03
openstackgerritFernando Diaz proposed openstack/castellan: Add created property to Managed Objects  https://review.openstack.org/23815017:06
*** stanzi has quit IRC17:06
*** Stanzi_ has quit IRC17:07
elmikodiazjf: did you work on the castellan.common.config module?17:07
diazjfhey elmiko, yeah I added this patch: https://review.openstack.org/#/c/238328/17:10
elmikodiazjf: ok, cool. i'm pretty sure it's causing a bug for downstream consumers =(17:13
elmikoi added, https://bugs.launchpad.net/castellan/+bug/152126517:13
openstackLaunchpad bug 1521265 in castellan "Log configuration causes error in applications that have already processed configuration arguments" [Undecided,New]17:13
elmikoand an agenda item to talk about it today at the meeting17:13
elmikoi think we need to re-assess the log usage in castellan17:13
diazjfelmiko, It was just to allow output to stdout, but I want to in the future have logging to file support. I'll take a look at it17:14
diazjfthanks for bringing it up17:14
elmikodiazjf: unfortunately, i think we need to add it as an option. we can't bake it into the lib.17:14
elmikoideally, consumers of castellan will be configuring their own logging support. the lib should accept that usage, and only provide the option to create logging if the user explicitly asks for it.17:15
elmikoalso, logging to a file exists as a cli option in oslo.log (--log-file, iirc)17:16
*** Stanzi has joined #openstack-barbican17:17
diazjfelmiko, I see, so we should have stdout as optional if the user specifies it in the config.17:18
elmikodiazjf: well, we should default to however the user decides to configure their logging. but, i could see keeping that configuration function for the purpose of allowing a user to call it if they have no logging support in their apps.17:19
elmikoit's just that most openstack services, that might consume castellan, will already have configured their logging.17:20
*** jorge_munoz has quit IRC17:21
diazjfelmiko gotcha, that sounds like a good idea. Do you have the log_error_test.py17:22
diazjfI wanna recreate the error17:22
*** Stanzi has quit IRC17:22
diazjfand I'll work on a fix, for if the configuration is none then set it to stdout17:22
elmikosure thing, 1sec17:23
elmikodiazjf: http://paste.openstack.org/show/480403/17:24
elmikodiazjf: i think any solution should be an optional call instead of occurring during the API call17:25
elmikothe reason i added the configuration that occurs currently during the API() call is that those functions are re-entrant safe.17:25
*** gyee has joined #openstack-barbican17:28
*** stanzi has joined #openstack-barbican17:33
*** Stanzi_ has joined #openstack-barbican17:33
diazjfelmiko, a quick fix would be http://paste.openstack.org/show/480406/ making it the default if it hasn't been set. Any suggestions for making it an optional call.17:35
elmikodiazjf: i don't think that would fix it17:37
elmikoas for making it optional, just remove the call from API()17:37
elmikooh, and i added a note to the bug, but the configure_logging call needs to accept a ConfigOpt object instead of using the global cfg.CONF object17:37
elmikothe rest of the configuration options in castellan are setup to allow the user to define their own config object, this should follow the same model. just in case the user isn't using the global CONF for their configuration.17:38
*** stanzi has quit IRC17:39
*** Stanzi_ has quit IRC17:40
diazjfelmkio, gotcha I look further into this. I did run python log_error_test.py with the change posted and no errors were returned. thanks for all the details I appreciate it17:41
elmikodiazjf: right, because in the patch you posted it will never get called :/17:42
elmikoconf will always be set to something17:43
diazjflol ;)17:43
elmiko;)17:43
elmikohonestly, we should discuss more at the meeting today. i'm curious to hear what the group has to say in general about logging for castellan17:44
diazjfelmkio, thanks, yeah I'll continue to look at it till then.17:44
elmikoi am looking through some other libraries to get a wider view of how this is handled17:44
elmikodiazjf: thanks!17:44
diazjfno problem17:45
*** lisaclark1 has quit IRC17:45
*** pdesai has joined #openstack-barbican17:47
*** lisaclark1 has joined #openstack-barbican17:54
*** dimtruck is now known as zz_dimtruck17:55
*** kfarr has joined #openstack-barbican18:02
*** zz_dimtruck is now known as dimtruck18:02
*** Kevin_Zheng has quit IRC18:06
*** stanzi has joined #openstack-barbican18:06
*** chlong has quit IRC18:10
*** stanzi has quit IRC18:11
*** silos has quit IRC18:16
*** rm_work has quit IRC18:17
*** Stanzi has joined #openstack-barbican18:22
*** dave-mccowan has quit IRC18:22
*** jillysciarilly has quit IRC18:22
*** rm_work has joined #openstack-barbican18:22
*** reaperhulk has quit IRC18:23
*** alee has quit IRC18:23
*** chlong has joined #openstack-barbican18:26
*** Stanzi has quit IRC18:27
*** jillysciarilly has joined #openstack-barbican18:28
*** dave-mccowan has joined #openstack-barbican18:32
*** stanzi has joined #openstack-barbican18:38
*** stanzi has quit IRC18:42
*** alee has joined #openstack-barbican18:45
*** Stanzi has joined #openstack-barbican18:53
*** Stanzi has quit IRC18:58
*** jhfeng has joined #openstack-barbican19:07
*** jaosorior has joined #openstack-barbican19:08
*** stanzi has joined #openstack-barbican19:09
*** Stanzi_ has joined #openstack-barbican19:09
*** Stanzi_ has quit IRC19:14
*** stanzi has quit IRC19:14
*** jaosorior has quit IRC19:15
*** reaperhulk has joined #openstack-barbican19:18
*** xek has quit IRC19:21
*** Stanzi_ has joined #openstack-barbican19:25
*** stanzi has joined #openstack-barbican19:25
*** silos has joined #openstack-barbican19:26
*** lisaclark1 has quit IRC19:27
*** lisaclark1 has joined #openstack-barbican19:28
*** lisaclark2 has joined #openstack-barbican19:29
*** lisaclark1 has quit IRC19:30
*** Stanzi_ has quit IRC19:30
*** stanzi has quit IRC19:30
*** lisaclark2 has quit IRC19:31
*** lisaclark1 has joined #openstack-barbican19:31
*** Stanzi has joined #openstack-barbican19:41
*** lisaclark1 has quit IRC19:43
*** Stanzi has quit IRC19:45
diazjfelmiko, just updated the bug with some thoughts19:50
elmikodiazjf: ack, i'll take a gander19:50
diazjfelmiko, coolio19:51
diazjfkfarr, I don't think Nate is able to join, do you wanna lead the meeting, if not I'd be happy to do it :)19:53
elmikodiazjf: i don't think that patch will completely solve my issues, i'll make a detailed note in the bug19:54
kfarrdiazjf, I was just thinking that19:54
diazjfelmiko, gotcha we'll continue to hack at it19:54
diazjfkfarr, I wrote down all the commands for it lol19:55
kfarrdiazjf nice!19:55
*** dimtruck is now known as zz_dimtruck19:56
*** Stanzi_ has joined #openstack-barbican19:56
*** stanzi has joined #openstack-barbican19:56
diazjfMeeting Starting in 2 min19:58
kfarrdiazjf, you or me?  I can do it!19:59
*** edtubill has joined #openstack-barbican19:59
diazjfkfarr, You can do it if you want if not I gotcha covered19:59
diazjfkfarr, you beat me too it :p20:00
diazjflol20:00
kfarrmeeting starting now in #openstack-meeting-alt !20:00
*** Stanzi_ has quit IRC20:01
*** stanzi has quit IRC20:01
*** lisaclark1 has joined #openstack-barbican20:11
*** zz_dimtruck is now known as dimtruck20:11
*** stanzi_ has joined #openstack-barbican20:12
*** Stanzi has joined #openstack-barbican20:12
*** stanzi_ has quit IRC20:17
*** Stanzi has quit IRC20:17
*** lisaclark1 has quit IRC21:05
kfarrhey diazjf thanks for offering to lead the meeting today! that was very helpful21:10
*** lisaclark1 has joined #openstack-barbican21:10
diazjfkfarr, no worries anytime!21:10
*** lisaclark1 has quit IRC21:12
diazjfelmiko, just one comment on https://bugs.launchpad.net/castellan/+bug/1521265 and I'll get a patch up21:12
openstackLaunchpad bug 1521265 in castellan "Log configuration causes error in applications that have already processed configuration arguments" [Undecided,New] - Assigned to Fernando Diaz (diazjf)21:12
elmikodiazjf: k21:14
*** Stanzi_ has joined #openstack-barbican21:15
*** stanzi has joined #openstack-barbican21:15
elmikodiazjf: answered, but i think there are some deeper questions surrounding proper usage of the library.21:18
diazjfelmiko, I'll take a look21:19
diazjfthanks21:19
elmikonp, i'll be around for a little while longer, so if you want to talk it out just give me a ping21:19
*** Stanzi_ has quit IRC21:20
*** stanzi has quit IRC21:20
diazjfelmiko, sure, lets talk a little more about it21:20
*** Stanzi has joined #openstack-barbican21:22
*** stanzi_ has joined #openstack-barbican21:22
elmikodiazjf: cool21:23
diazjfelmiko, so in the key_manager I just want enable_logging(configuration=None) so that the user can add a conf or if none is passed then we will use default logging21:23
diazjfIn config.py we have default values setup which are setup only if no conf is passed21:24
elmikodiazjf: i don't think you need to enable logging in the key_manager21:25
elmikoimo, i think this patch http://paste.openstack.org/show/480430/ would do the trick21:27
elmikobut, it leaves the configure_logging option up to the user21:27
diazjfelmiko, so the user will have to import config and then run configure_logging?21:28
elmikodiazjf: yea, something like that21:28
elmikoand then they could run that configure_logging at an appropriate time in their application21:28
diazjfelmiko, sounds good, We will just have to add some good docs21:28
elmikodiazjf: yes, +121:28
diazjfelmiko, awesome I'll add you as a co-author21:29
elmikoand really, that should only be used if they have not already configured logging21:29
elmikok21:29
diazjfelmiko, or if you already have the patch its cool :)21:29
elmikodiazjf: no, go for it!21:30
*** lisaclark1 has joined #openstack-barbican21:31
elmikodiazjf: this will fix it for now, then we can work on something more robust for the future. i think there is probably a better way we can handle this, for example with some castellan specific configuration options.21:32
*** stanzi_ has quit IRC21:33
*** Stanzi has quit IRC21:33
diazjfelmiko, thats a good idea21:34
*** kfarr has quit IRC21:34
elmikodiazjf: just thinking here, should we put this function in castellan.options with the stuff to set defaults and list options?21:35
elmikosince those are already documented as an entry point into castellan configurations?21:36
diazjfelmiko, I think so, better not to have settings everywhere21:37
elmikodiazjf: that's kinda what i was thinking too21:38
diazjfI created http://paste.openstack.org/show/480435/21:39
diazjfbut does not hold changes when called after API21:39
*** kfarr has joined #openstack-barbican21:39
elmikodiazjf: when did you call enable_logging?21:40
diazjfhttp://paste.openstack.org/show/480436/21:40
diazjfI also added an log.error in the store key function21:41
elmikook, sounds like you might need to add the log.setup in there, but i would do something like:21:41
elmikodef enable_logging(conf=None, appname='castellan')21:41
elmikothen just use21:41
elmikolog.setup(conf, appname)21:42
elmikoso that the application name can be overridden by the end user if necessary21:42
diazjfelmiko, alright let me give it a shot21:43
elmikocool21:44
elmikodiazjf: you may also need to call cfg.CONF() from your app, to process the configuration options21:50
elmikothis is kinda why i think it's a blurring of lines to have the library do some of this work for the user21:51
diazjfelmiko, http://paste.openstack.org/show/480438/ is working for me21:53
diazjfjust can't register_options after API is called so I shouldn't include it21:53
diazjfI'm running http://paste.openstack.org/show/480439/21:53
diazjfSo I can remove castellan.common.config and add this code to castellan.options21:54
elmikodiazjf: if we're gonna contain all the logging stuff in that enable_logging call, i think it would be better to add the log.register_conf in that function and just tell people to run it before calling API()21:56
elmikoit seems weird to setup logging inside enable_logging, then require an extra step from the user with regards to the log21:56
elmikolike, ideally, you should just have the call to log.register_conf inside enable_logging, then call enable_logging() right before you do cfg.CONF()21:57
diazjfelmiko, thats true, a user should know that they are setting up beforehand.21:58
elmikoright21:59
elmikoand they elect not to do any logging stuff, then they can call this convenience function21:59
elmikootherwise, logging will work however they configure it21:59
diazjfelmiko, I'll have a patch up in 1522:03
elmikodiazjf: awesome22:04
*** notmyname has quit IRC22:05
*** notmyname has joined #openstack-barbican22:06
*** spotz is now known as spotz_zzz22:14
*** spotz_zzz is now known as spotz22:19
openstackgerritFernando Diaz proposed openstack/castellan: Allow for default logging configuration to be user enabled  https://review.openstack.org/25155822:19
diazjfelmiko ^22:20
elmikodiazjf: ack, taking a look22:20
diazjfelmiko, cool lmk and I'll start on the docs22:23
elmikohehe, was just about to comment about the docs22:26
elmikodiazjf: ok, added my nits22:26
diazjfelmiko, awesome thanks. I'll take a look and add some docs later tonight22:27
*** silos has left #openstack-barbican22:29
elmikodiazjf: cool22:31
openstackgerritFernando Diaz proposed openstack/castellan: Allow for default logging configuration to be user enabled  https://review.openstack.org/25155822:35
*** jhfeng has quit IRC22:37
openstackgerritFernando Diaz proposed openstack/castellan: Allow for default logging configuration to be user enabled  https://review.openstack.org/25155822:39
*** jhfeng has joined #openstack-barbican22:47
*** Stanzi has joined #openstack-barbican22:51
*** stanzi_ has joined #openstack-barbican22:51
*** diazjf has quit IRC22:54
*** stanzi_ has quit IRC22:56
*** Stanzi has quit IRC22:56
*** jhfeng has quit IRC22:56
*** kfarr has left #openstack-barbican22:56
*** lisaclark1 has quit IRC23:02
*** lisaclark1 has joined #openstack-barbican23:05
*** Stanzi has joined #openstack-barbican23:07
*** spotz is now known as spotz_zzz23:08
*** edtubill has quit IRC23:09
*** Stanzi has quit IRC23:11
*** Stanzi has joined #openstack-barbican23:23
*** jmckind has quit IRC23:23
*** stanzi_ has joined #openstack-barbican23:23
*** lisaclark1 has quit IRC23:26
*** spotz_zzz is now known as spotz23:26
*** lisaclark1 has joined #openstack-barbican23:26
*** stanzi_ has quit IRC23:27
*** Stanzi has quit IRC23:27
*** lisaclark1 has quit IRC23:31
*** pdesai has quit IRC23:35
*** barra204 has quit IRC23:38
*** shakamunyi has joined #openstack-barbican23:38
*** Stanzi has joined #openstack-barbican23:38
*** pdesai has joined #openstack-barbican23:40
openstackgerritAdam Harwell proposed openstack/barbican: Remove consumer check for project_id to match containers  https://review.openstack.org/25116823:40
rm_workdave-mccowan: ^^ I think that should work? we'll see when the gate runs23:40
rm_workdave-mccowan: your comment "Also, the project id still needs to be saved in the database to support the quota feature." I am not sure if that is a problem23:41
rm_workdave-mccowan: it should still save, the line i removed was a duplicate I believe, so seems like I removed it but it should be in there from the constructor23:41
*** Stanzi has quit IRC23:43
*** spotz is now known as spotz_zzz23:49
*** dave-mccowan has quit IRC23:51
*** dave-mccowan has joined #openstack-barbican23:58
« Saturday, 2015-11-28 Index Tuesday, 2015-12-01 »

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