| openstackgerrit | Edwin Zhai proposed openstack/watcher: Handle race condition of sync from multiple decision engines https://review.openstack.org/339285 | 00:08 |
|---|---|---|
| *** michaelgugino has quit IRC | 00:08 | |
| *** thorst has joined #openstack-watcher | 00:12 | |
| *** thorst has quit IRC | 00:17 | |
| *** thorst has joined #openstack-watcher | 00:53 | |
| *** thorst has quit IRC | 00:58 | |
| *** thorst has joined #openstack-watcher | 01:14 | |
| *** thorst has quit IRC | 01:19 | |
| *** esberglu has joined #openstack-watcher | 01:29 | |
| *** esberglu has quit IRC | 01:35 | |
| *** jwcroppe has joined #openstack-watcher | 01:45 | |
| openstackgerrit | Rong Han proposed openstack/watcher-dashboard: [WIP]Add policies for API access control to watcher-dashboard project https://review.openstack.org/339082 | 01:52 |
| *** jinquan has quit IRC | 01:54 | |
| *** jwcroppe has quit IRC | 01:57 | |
| openstackgerrit | Rong Han proposed openstack/watcher: Optimize local.conf.controller file to enable watche-dashboard https://review.openstack.org/339309 | 02:01 |
| *** hanrong has left #openstack-watcher | 02:05 | |
| *** thorst has joined #openstack-watcher | 02:25 | |
| openstackgerrit | Rong Han proposed openstack/watcher-dashboard: [WIP]Add policies for API access control to watcher-dashboard project https://review.openstack.org/339082 | 02:40 |
| openstackgerrit | Rong Han proposed openstack/watcher-dashboard: [WIP]Add policies for API access control to watcher-dashboard https://review.openstack.org/339082 | 02:41 |
| openstackgerrit | Rong Han proposed openstack/watcher-dashboard: Modify error message from 'action_plan' to 'audit' https://review.openstack.org/339319 | 02:45 |
| *** jwcroppe has joined #openstack-watcher | 02:57 | |
| *** jwcroppe has quit IRC | 03:01 | |
| *** jwcroppe has joined #openstack-watcher | 03:01 | |
| *** jwcroppe has quit IRC | 03:40 | |
| *** jwcroppe has joined #openstack-watcher | 03:51 | |
| *** jwcroppe has quit IRC | 03:55 | |
| *** jwcroppe has joined #openstack-watcher | 04:17 | |
| *** jwcroppe has quit IRC | 04:33 | |
| *** jwcroppe has joined #openstack-watcher | 05:13 | |
| *** jwcroppe has quit IRC | 05:49 | |
| *** jwcroppe has joined #openstack-watcher | 05:49 | |
| *** jwcroppe has quit IRC | 06:12 | |
| *** jwcroppe has joined #openstack-watcher | 06:13 | |
| *** danpawlik has joined #openstack-watcher | 06:15 | |
| *** jwcroppe has quit IRC | 06:42 | |
| *** jwcroppe has joined #openstack-watcher | 06:44 | |
| *** dtardivel has joined #openstack-watcher | 07:06 | |
| openstackgerrit | Tomasz Kaczynski proposed openstack/python-watcherclient: Add scoring engine commands https://review.openstack.org/330385 | 07:09 |
| *** diga has joined #openstack-watcher | 07:22 | |
| acabot | danpawlik : Hi, I saw https://review.openstack.org/#/c/338585/ | 07:34 |
| danpawlik | acabot: hello | 07:34 |
| danpawlik | acabot: should I add "missing" parameters or not ? | 07:35 |
| *** jwcroppe has quit IRC | 07:35 | |
| acabot | danpawlik : just to let you know that Watcher doc is now at http://docs.openstack.org/developer/watcher/ | 07:35 |
| danpawlik | acabot: oh, ok | 07:35 |
| acabot | danpawlink : so please do not refer to link on factory.b-com.com (we will shut it down soon) ;-) | 07:35 |
| danpawlik | acabot: ok ok :) | 07:36 |
| acabot | danpawlik : dtardivel will check "missing" parameters | 07:36 |
| danpawlik | acabot: ok, so I'm waiting for a response :) | 07:37 |
| *** jwcroppe has joined #openstack-watcher | 07:38 | |
| acabot | danpawlik : by the way, how did you get this link ? https://factory.b-com.com/www/watcher/doc/watcher/deploy/conf-files.html ? | 07:39 |
| acabot | danpawlik : I'm trying to check where we still have links on factory.b-com.com | 07:40 |
| danpawlik | acabot: some times ago someone paste it on #openstack-watcher | 07:40 |
| acabot | danpawlik : ok cool :-) thx | 07:40 |
| danpawlik | acabot: second posision on google is URL to b-com: watcher conf sample | 07:41 |
| danpawlik | acabot: it will be hard to remove URL from google xD | 07:42 |
| acabot | search query ? | 07:42 |
| danpawlik | acabot: watcher conf sample | 07:42 |
| acabot | my second link is docs.openstack.org/developer/watcher/deploy/configuration.html | 07:43 |
| acabot | 07:43 | |
| dtardivel | hi danpawlik: did you read this documentation section: https://factory.b-com.com/www/watcher/doc/watcher/deploy/configuration.html#configure-the-watcher-service | 07:43 |
| acabot | google poland ? | 07:43 |
| dtardivel | we explain how to configure [keystone_authtoken] section with domain parameter | 07:44 |
| danpawlik | acabot: http://imgur.com/a/MYazG | 07:44 |
| danpawlik | acabot: I think google is "caching" my history | 07:45 |
| acabot | danpawlik : thats weird, factory has a robots.txt file | 07:45 |
| acabot | danpawlik : can you try with a "clean" browser ? | 07:45 |
| danpawlik | dtardivel: hmm, I modeled puppet-watcher module on watcher.conf.sample file. | 07:46 |
| danpawlik | acabot: sure | 07:47 |
| danpawlik | acabot: in private browsering mode, I don't see any URL from b-com :) | 07:48 |
| danpawlik | acabot: but on my account after clearing cache, one more time sign it on google accout -> and b-com is still second :D | 07:48 |
| danpawlik | google <3 | 07:48 |
| *** alexchadin has joined #openstack-watcher | 07:53 | |
| *** jinquan has joined #openstack-watcher | 07:58 | |
| *** alexchadin has quit IRC | 08:08 | |
| *** alexchadin has joined #openstack-watcher | 08:09 | |
| alexchadin | hi dtardivel | 08:13 |
| *** jinquan has quit IRC | 08:16 | |
| *** jinquan has joined #openstack-watcher | 08:16 | |
| danpawlik | dtardivel: in documentation you have user_domain_id, but it's deprecated :P | 08:21 |
| alexchadin | acabot: hi | 08:25 |
| alexchadin | about resubmitting | 08:25 |
| openstackgerrit | Jinquan Ni proposed openstack/watcher: Make default Planner generic to handle new action https://review.openstack.org/338921 | 08:26 |
| alexchadin | it is going through git-review -d ? | 08:26 |
| *** hanrong has joined #openstack-watcher | 08:32 | |
| alexchadin | acabot: are you here? | 08:33 |
| *** jwcroppe has quit IRC | 08:34 | |
| openstackgerrit | Athlan-Guyot sofer proposed openstack/puppet-watcher: Include openstacklib defaults manifest. https://review.openstack.org/338848 | 08:54 |
| vincentfrancoise | alexchadin: acabot's not at his desk right now | 08:55 |
| openstackgerrit | Rong Han proposed openstack/watcher-dashboard: [WIP]Add policies for API access control to watcher-dashboard https://review.openstack.org/339082 | 08:55 |
| *** hanrong1 has joined #openstack-watcher | 08:55 | |
| *** jinquan1 has joined #openstack-watcher | 08:56 | |
| alexchadin | vincentfrancoise: seems that define audit scope is in dependency with your CDMC patches | 08:57 |
| *** jinquan has quit IRC | 08:58 | |
| *** hanrong has quit IRC | 08:58 | |
| vincentfrancoise | if you think so then we should update the dependency graph on launchpad | 08:58 |
| danpawlik | dtardivel: I will add into puppet-module all keystonemiddleware parameters | 09:00 |
| vincentfrancoise | but that doesn't suprise me since the audit scoping is directly linked to the cluster models | 09:00 |
| *** alexchadin has quit IRC | 09:01 | |
| *** alexchadin has joined #openstack-watcher | 09:01 | |
| *** alexchadin has quit IRC | 09:31 | |
| *** alexchadin has joined #openstack-watcher | 09:31 | |
| danpawlik | dtardivel: do you agree :> ? | 09:47 |
| vincentfrancoise | danpawlik: he's not at his desk | 09:47 |
| vincentfrancoise | will probably get back after lunch break | 09:47 |
| danpawlik | vincentfrancoise: so I will push my PS | 09:48 |
| *** alexchad_ has joined #openstack-watcher | 09:48 | |
| *** alexchadin has quit IRC | 09:48 | |
| openstackgerrit | Jinquan Ni proposed openstack/watcher: Make default Planner generic to handle new action https://review.openstack.org/338921 | 09:49 |
| vincentfrancoise | I guess you can | 09:49 |
| danpawlik | vincentfrancoise: if we use keystonemiddleware lib but some parameters are missing in generated watcher sample file (like check_revocations_for_cached, enforce_token_bind, etc) these parameters should be "enabled" in watcher or ? | 09:52 |
| vincentfrancoise | hmm, the whole keystone subject is a massive headache (at least last time I checked...) | 09:52 |
| danpawlik | vincentfrancoise: i know xD | 09:52 |
| *** alexchad_ has quit IRC | 09:53 | |
| *** jwcroppe has joined #openstack-watcher | 09:53 | |
| vincentfrancoise | I don't want to tell you something wrong, that's why I was waiting for dtardivel which knows more about these since he internally re-deployed watcher recently | 09:54 |
| *** diga has quit IRC | 09:54 | |
| openstackgerrit | Daniel Pawlik proposed openstack/puppet-watcher: Fix authtoken options and adding the missing https://review.openstack.org/338585 | 09:56 |
| vincentfrancoise | as far as I can see, all parameters in https://github.com/openstack/keystonemiddleware/blob/master/keystonemiddleware/auth_token/_opts.py should be part of the generated config sample | 09:58 |
| vincentfrancoise | and it includes enforce_token_bind and check_revocations_for_cached | 09:58 |
| vincentfrancoise | so if you don't see it that's odd | 09:59 |
| *** diga has joined #openstack-watcher | 10:01 | |
| vincentfrancoise | I just generated the sample file locally I can see check_revocations_for_cached and enforce_token_bind in it | 10:01 |
| danpawlik | yap, I see | 10:02 |
| danpawlik | I was looking at this config file from b-com | 10:02 |
| danpawlik | So I should change commit message xD | 10:02 |
| *** alexchadin has joined #openstack-watcher | 10:02 | |
| vincentfrancoise | they should be enabled in watcher thanks to https://github.com/openstack/watcher/blob/master/watcher/api/hooks.py#L64 | 10:03 |
| danpawlik | yeah, I see that they are included in watcher ;) | 10:03 |
| vincentfrancoise | so I didn't test them but I fairly confident they are taken into account | 10:04 |
| danpawlik | vincentfrancoise: I will compare all puppet-watcher parameters with new generated watcher config file and then I fixed it. For now I will push "all manifests" | 10:07 |
| vincentfrancoise | sounds good to me :) | 10:08 |
| vincentfrancoise | off for lunch see you later | 10:08 |
| danpawlik | vincentfrancoise: bon appétit | 10:11 |
| *** jwcroppe has quit IRC | 10:17 | |
| openstackgerrit | Alexander Chadin proposed openstack/watcher-specs: Update define-the-audit-scope spec https://review.openstack.org/339456 | 10:37 |
| *** alexchadin has quit IRC | 10:41 | |
| openstackgerrit | Rong Han proposed openstack/watcher-dashboard: [WIP]Add policies for API access control to watcher-dashboard https://review.openstack.org/339082 | 10:45 |
| *** diga has quit IRC | 11:01 | |
| *** jwcroppe has joined #openstack-watcher | 11:02 | |
| *** hanrong1 has left #openstack-watcher | 11:04 | |
| *** jwcroppe has quit IRC | 11:09 | |
| *** jinquan has joined #openstack-watcher | 11:11 | |
| openstackgerrit | Jinquan Ni proposed openstack/watcher: Make default Planner generic to handle new action https://review.openstack.org/338921 | 11:31 |
| dtardivel | danpawlik: Some authtoken parameters are not referenced in the built watcher.conf.sample, because they are auth plugin parameters (it is a bizarrerie from keystone libs) . This is the reason why we added a dedicated configuration section in the Watcher doc explaining how to use authtoken middleware with the 'password' plugin and its related additional | 11:35 |
| dtardivel | parameters. | 11:35 |
| danpawlik | dtardivel: ok, thx :) | 11:36 |
| *** jwcroppe has joined #openstack-watcher | 11:39 | |
| *** jwcroppe has quit IRC | 11:41 | |
| *** thorst has joined #openstack-watcher | 11:48 | |
| tkaczynski | hi. I'm trying to create a new audit with watcher CLI. it requires to provide "period", which is not documented in watcher user guide. what value should I use there? | 12:25 |
| vincentfrancoise | hi tkaczynski | 12:26 |
| vincentfrancoise | are you on the master branch? | 12:26 |
| tkaczynski | kind of. I'm on my bp/scoring-engine branch, which is based on master, but possibly from few weeks back | 12:27 |
| vincentfrancoise | because period should have been renamed interval | 12:27 |
| tkaczynski | well, it's not then. but it doesn't matter I guess? the value should be the same? | 12:28 |
| vincentfrancoise | for a couple of days we merged the CLI that set it as period although it was a mistake | 12:28 |
| vincentfrancoise | and it was rolled back | 12:28 |
| tkaczynski | so it won't work for me if I use period? | 12:29 |
| vincentfrancoise | so if you rebase onto the latest version of master it should disappear I believe | 12:29 |
| tkaczynski | but the interval will still be required? | 12:29 |
| vincentfrancoise | it should have a default that makes it kind of optional I think | 12:30 |
| tkaczynski | ok. do I need to rebase watcher project as well? or just the CLI? | 12:31 |
| vincentfrancoise | the the CLI | 12:31 |
| vincentfrancoise | just* | 12:31 |
| tkaczynski | ok. then I guess I should upload the new patchset for the review... ? | 12:31 |
| tkaczynski | ok, now it cries for audit type, which is also not documented ;) | 12:34 |
| vincentfrancoise | then you need to rebase your watcher this time | 12:35 |
| vincentfrancoise | and recreate the watcher DB | 12:35 |
| vincentfrancoise | :p | 12:36 |
| tkaczynski | awesome ;) | 12:36 |
| vincentfrancoise | we're merging lots of changesets these days so you'll have to rebase quite often | 12:37 |
| sballe_ | morning | 12:37 |
| sballe_ | acabot: I cannot call you it says the number is not in service | 12:37 |
| sballe_ | Can you call me? | 12:37 |
| tkaczynski | this is exactly I'm hoping to get my change merged too | 12:37 |
| vincentfrancoise | I believe I +2'd your code | 12:38 |
| tkaczynski | but ideally watcher should be working all the time, no? | 12:38 |
| sballe_ | tkaczynski: yes in therory | 12:38 |
| vincentfrancoise | but it won't get merged before we tag the n2 release | 12:38 |
| tkaczynski | vincentfrancoise: yes, you did. can't wait :) | 12:39 |
| vincentfrancoise | yes the CLI merge was a mistake that wasn't vaught because we lack integration tests on the CLI | 12:39 |
| vincentfrancoise | cross repo failure | 12:39 |
| tkaczynski | understand. will rebase and restack | 12:40 |
| tkaczynski | vincentfrancoise: as always, thanks for help :) | 12:40 |
| vincentfrancoise | you're welcome, as always :) | 12:42 |
| *** edleafe is now known as figleaf | 12:46 | |
| dtardivel | tkaczynski: you issue has been fixed in the changeset https://review.openstack.org/#/c/339026/5 | 12:46 |
| dtardivel | it will be merged asap this afternoon | 12:47 |
| tkaczynski | dtardivel: thanks for the update :) | 12:48 |
| *** bauzas has joined #openstack-watcher | 12:48 | |
| tkaczynski | one more rebase then ;) | 12:48 |
| dtardivel | tkaczynski: thks to Vladimir .... | 12:48 |
| *** acabot_ has joined #openstack-watcher | 12:51 | |
| *** alexchadin has joined #openstack-watcher | 13:05 | |
| *** acabot_ has quit IRC | 13:08 | |
| *** acabot_ has joined #openstack-watcher | 13:10 | |
| *** dronshaw has joined #openstack-watcher | 13:12 | |
| tkaczynski | vincentfrancoise: any particular reason that I need to recreate watcher DB? is it not enough to re-stack / recreate tables? | 13:14 |
| vincentfrancoise | tkaczynski: it's just easier to drop the DB, recreate the DB ad play the watcher-db-manage create_schema on top of it | 13:15 |
| vincentfrancoise | but if you have another way to do it that is faster, then feel free | 13:16 |
| tkaczynski | I'm asking because after the rebase and unstack/stack (which is deleting all watcher tables and creating them again) the watcher has failed to start (causing the whole devstack to fail) | 13:17 |
| vincentfrancoise | to me devstack is a wild beast that you shouldn't poke around unless you really have to | 13:21 |
| vincentfrancoise | so I avoid using ./unstack as much as I can | 13:21 |
| vincentfrancoise | if I want to update watcher to test it | 13:22 |
| vincentfrancoise | then I kill ctrl+C the process in screen | 13:22 |
| vincentfrancoise | and restart it after appying the changes in /opt/stack/watcher | 13:23 |
| tkaczynski | sure, I do the same. but for things like db-schema update I'm not so sure which process is responsible for DB schema creation etc. maybe even it's part of the devstack setup, not the particular service | 13:23 |
| vincentfrancoise | oh there is none | 13:24 |
| tkaczynski | so to be safe I usually do restack | 13:24 |
| vincentfrancoise | I mean | 13:24 |
| vincentfrancoise | I literally do | 13:24 |
| vincentfrancoise | mysql -uroot -pstackdb | 13:24 |
| vincentfrancoise | DROP DATABASE watcher; | 13:24 |
| vincentfrancoise | CREATE DATABASE watcher CHARACTER SET utf8; | 13:24 |
| *** Zucan has joined #openstack-watcher | 13:24 | |
| vincentfrancoise | then | 13:24 |
| vincentfrancoise | watcher-db-manage create_schema | 13:25 |
| vincentfrancoise | DOne | 13:25 |
| tkaczynski | so exactly this command has failed for me :( | 13:25 |
| vincentfrancoise | oh | 13:26 |
| vincentfrancoise | what's the failure then | 13:26 |
| vincentfrancoise | ? | 13:26 |
| tkaczynski | https://www.irccloud.com/pastebin/FRTfuNBY/ | 13:26 |
| alexchadin | tkaczynski: have you tried to specify --config-file option? | 13:27 |
| vincentfrancoise | look like a bug you found out here | 13:27 |
| vincentfrancoise | coming from the policy BP | 13:27 |
| tkaczynski | the --config-file option is specified as you see | 13:29 |
| tkaczynski | so what's this bug then? looks I just lost my environment :( | 13:29 |
| vincentfrancoise | doesn't change anything | 13:29 |
| vincentfrancoise | by default --config-file looks for a /etc/watcher/watcher.conf anyway unless you override it | 13:29 |
| vincentfrancoise | tkaczynski: no if the bug gets fixed you should be good to go after a ./stack.sh | 13:30 |
| tkaczynski | sure, but who knows when the bug will be fixed. can I just remove this argument? | 13:31 |
| tkaczynski | (for the moment as a work-around) | 13:31 |
| vincentfrancoise | well | 13:31 |
| vincentfrancoise | let me give you my version of context.py onto which I currently work | 13:31 |
| vincentfrancoise | to see if it gets any better | 13:32 |
| tkaczynski | so the watcher is currently broken on master, right? | 13:33 |
| vincentfrancoise | yes | 13:33 |
| tkaczynski | awesome | 13:33 |
| vincentfrancoise | can you update you watcher/common/context.py to with http://pastebin.com/V9XptsA0 to see if it works? | 13:34 |
| openstackgerrit | Daniel Pawlik proposed openstack/puppet-watcher: Fix authtoken options and adding the missing https://review.openstack.org/338585 | 13:36 |
| *** esberglu has joined #openstack-watcher | 13:37 | |
| tkaczynski | I will, hang on | 13:37 |
| danpawlik | vincentfrancoise: http://paste.openstack.org/ is better as pastebin :> | 13:37 |
| danpawlik | than* | 13:37 |
| vincentfrancoise | danpawlik: thanks! | 13:37 |
| tkaczynski | same error, in different line though ;) | 13:40 |
| vincentfrancoise | -__- | 13:40 |
| vincentfrancoise | can you pastebin it as well? | 13:40 |
| tkaczynski | looks like oslo.context.__init__ doesn't accept roles argument | 13:40 |
| tkaczynski | dependency issue? | 13:41 |
| dtardivel | alexchadin: hi, is Vladimir available today ? | 13:41 |
| vincentfrancoise | https://github.com/openstack/oslo.context/blob/master/oslo_context/context.py#L72 | 13:41 |
| tkaczynski | https://www.irccloud.com/pastebin/7icf0RWH/ | 13:41 |
| vincentfrancoise | roles is there... | 13:41 |
| alexchadin | dtardivel: hi, he is | 13:42 |
| vincentfrancoise | tkaczynski: can you pastebin me the output of the 'pip list' command within your devstack please? | 13:43 |
| tkaczynski | I did run the watcher-db-schema manually now, without running the whole stask.sh | 13:43 |
| *** ostroverkhov has joined #openstack-watcher | 13:43 | |
| tkaczynski | oslo.context (2.1.0) | 13:44 |
| dtardivel | ok, I have a request for him about watcher dashboard. It would be nice to add in audit table the audit-type column because we will have now ONE_SHOT and CONTINOUS audit. | 13:44 |
| alexchadin | dtardivel: he is here now | 13:44 |
| dtardivel | Could he propose a new patchset on this changeset https://review.openstack.org/#/c/327095/ ? | 13:44 |
| alexchadin | dtardivel: I think so | 13:45 |
| alexchadin | let us discuss it | 13:45 |
| tkaczynski | vincentfrancoise: oslo.context master is on 2.5.0, I have 2.1.0 | 13:45 |
| vincentfrancoise | tkaczynski: https://github.com/openstack/watcher/blob/master/requirements.txt#L12 | 13:45 |
| vincentfrancoise | watcher expects oslo.context>=2.4.0 | 13:45 |
| vincentfrancoise | so that's the issue here | 13:46 |
| vincentfrancoise | so master is not broken | 13:46 |
| tkaczynski | so why I have old version then? | 13:46 |
| vincentfrancoise | dunno | 13:46 |
| vincentfrancoise | run a pip install --upgrade -r requirements.txt from /opt/stack/watcher | 13:47 |
| tkaczynski | sure, but isn't stack.sh supposed to do that anyway? | 13:47 |
| vincentfrancoise | told you that devstack is a wild beast | 13:48 |
| danpawlik | xD | 13:48 |
| dtardivel | https://www.irccloud.com/pastebin/azPb1Q3s/%20%20%20%20audit_type%20%3D%20horizon.tables.Column(%20%20%20%20%20%20%20%20%20'audit_type'%2C%20%20%20%20%20%20%20%20%20verbose_name%3D_('Audit%20Type')) | 13:49 |
| dtardivel | alexchadin: ostroverkhov : solution ^^^^ :) | 13:50 |
| tkaczynski | but we are humans. we have brain and should be able to defeat wild beasts :) | 13:50 |
| *** ostrover_ has joined #openstack-watcher | 13:50 | |
| vincentfrancoise | true but sometimes it'd be nice to stop the fightings | 13:52 |
| tkaczynski | ok, manual execution od watcher-db-manage tool worked. restarting devstack now | 13:53 |
| *** ostroverkhov has quit IRC | 13:53 | |
| alexchadin | dtardivel: about continuous - it would be great to store our audit jobs | 13:54 |
| alexchadin | dtardivel: I will integrate it in the second iteration | 13:54 |
| *** danpawlik has left #openstack-watcher | 13:56 | |
| alexchadin | dtardivel: I think we can merge it as is | 13:56 |
| *** danpawlik has joined #openstack-watcher | 13:56 | |
| alexchadin | dtardivel: and give more features in the future | 13:56 |
| vincentfrancoise | tkaczynski: good :) | 13:56 |
| alexchadin | vincentfrancoise: I have bad feelings about devstack | 13:57 |
| dtardivel | alexchadin: OK. Is it ok with Edwin comenta ? | 13:57 |
| alexchadin | vincentfrancoise: that's why I spend 2 days to install openstack from the sources and configure it | 13:58 |
| *** acabot_ has quit IRC | 13:58 | |
| vincentfrancoise | alexchadin: who doesn't? | 13:58 |
| tkaczynski | alexchadin: long time ago I spent like 3 weeks or something to configure devstack. the last thing I want to do is to try that again | 13:59 |
| alexchadin | dtardivel: yes, I answered him that his suggestions are looking good and I think of this in the future | 13:59 |
| alexchadin | tkaczynski: oh... | 13:59 |
| tkaczynski | it plays very badly with Intel corporate proxy | 14:00 |
| vincentfrancoise | oh... | 14:00 |
| dtardivel | alexchadin: ok. I merge "continuous audit" (watcher + CLI) and wait for last patchser for Dashboard. Good Work Guys Thks a lot :) | 14:00 |
| *** acabot_ has joined #openstack-watcher | 14:01 | |
| alexchadin | vincentfrancoise: we are using devstack only to take it to the students. Once tasks are done - devstack is unstacked ;D | 14:01 |
| alexchadin | tkaczynski: it was hard to me to get how to install it | 14:02 |
| vincentfrancoise | as for watcher, it's enough for most dev tasks if you manage to make it work | 14:02 |
| alexchadin | vincentfrancoise: absolutely agree | 14:02 |
| alexchadin | vincentfrancoise: and you always have latest code version | 14:03 |
| vincentfrancoise | alexchadin: yep | 14:03 |
| tkaczynski | alexchadin: it's hard for everyone I think. I managed to get some help with that process, but I don't think I could repeat it myself. so I babysit my devstack more than I do my real kids ;) | 14:04 |
| *** danpawlik has quit IRC | 14:04 | |
| vincentfrancoise | xD | 14:04 |
| alexchadin | tkaczynski: wow :D | 14:04 |
| *** Zucan has quit IRC | 14:05 | |
| openstackgerrit | Merged openstack/watcher: Fix watcher doc build error https://review.openstack.org/338005 | 14:07 |
| openstackgerrit | Merged openstack/watcher: Add continuously optimization https://review.openstack.org/321411 | 14:07 |
| alexchadin | vincentfrancoise: I will review some part of your code this weekend and waiting your review regarding audit scope spec :) | 14:08 |
| vincentfrancoise | alexchadin: tanks a lot! I appreciate the motivation ;) | 14:09 |
| openstackgerrit | David TARDIVEL proposed openstack/watcher-dashboard: Initializes EfficacyIndicator attributes https://review.openstack.org/339573 | 14:11 |
| tkaczynski | ok, so it seems that devstack is reverting oslo.context library back to 2.1.0 and I have the same problem again :( | 14:11 |
| *** alexchadin has quit IRC | 14:13 | |
| vincentfrancoise | maybe you have an old version of devstack | 14:14 |
| tkaczynski | I do. because I did freeze it (disabled RECLONE). it's because I wanted to have repeatable process | 14:14 |
| vincentfrancoise | or you have forced a stable branch to , say, stable/liberty release which is expecting a oslo.context>=2.1.0 | 14:15 |
| tkaczynski | apparently it became too old now | 14:15 |
| vincentfrancoise | +1 | 14:15 |
| vincentfrancoise | coffee break, see you in a bit ;) | 14:16 |
| tkaczynski | weekend break, see you in a longer bit ;) | 14:16 |
| openstackgerrit | Vladimir Ostroverkhov proposed openstack/watcher-dashboard: Add support continuously-optimization https://review.openstack.org/327095 | 14:17 |
| *** diga has joined #openstack-watcher | 14:19 | |
| *** bauzas is now known as bauwser | 14:19 | |
| *** acabot_ has quit IRC | 14:32 | |
| *** acabot_ has joined #openstack-watcher | 14:32 | |
| *** acabot_ has quit IRC | 14:33 | |
| dtardivel | ostrover_: Vladimir, something is missing in the Dashboard. This is no way to stop a continuous audit. | 14:37 |
| ostrover_ | @dtardivel oh, It seems I forgot all about it, now I try to fix | 14:39 |
| dtardivel | and one question: Can we easily update the interval on a ongoing continuous audit ? | 14:40 |
| *** acabot_ has joined #openstack-watcher | 14:41 | |
| *** alexchadin has joined #openstack-watcher | 14:41 | |
| alexchadin | dtardivel: about updating audit interval. Sometimes we have a discussion with vincentfrancoise and decided that we can delete audit and create new instead of modifying interval of existing one. | 14:45 |
| dtardivel | alexchadin: vincentfrancoise for a first iteration, it's ok for me. | 14:46 |
| dtardivel | I have the solution for deleting the audit into dashboard. Do you want it ? | 14:46 |
| dtardivel | ostrover_: ^^ | 14:48 |
| vincentfrancoise | alexchadin: I just read the changes you made to the audit-scope spec and I'm sure I understand your code sample | 14:49 |
| vincentfrancoise | alexchadin: what is self.clients here https://review.openstack.org/#/c/339456/1/specs/newton/approved/define-the-audit-scope.rst@128 ? | 14:49 |
| alexchadin | dtardivel: at that date audit object doesn't have parameters field. Now when audit is getting more and more heavy we can implement updating of interval, I think | 14:50 |
| alexchadin | dtardivel: I plan to continue work on continuously-optimization | 14:50 |
| ostrover_ | dtardivel: I can look at it , if you want I can add it to commit | 14:51 |
| alexchadin | vincentfrancoise: just calling python client api | 14:51 |
| dtardivel | https://www.irccloud.com/pastebin/TFnNnHJQ/ | 14:52 |
| dtardivel | ostrover_: ^^ | 14:52 |
| dtardivel | ostrover_: tested in my devstack :) | 14:52 |
| alexchadin | vincentfrancoise: oh, seems that I was wrong with referencing | 14:52 |
| dtardivel | ostrover_: please add it in your changeset if it's ok for you | 14:53 |
| alexchadin | vincentfrancoise: It was rally client call convention | 14:53 |
| alexchadin | vincentfrancoise: If I remember right, Watcher has something like self.osc.nova() | 14:53 |
| ostrover_ | dtardivel: thank you, I'll look | 14:54 |
| vincentfrancoise | alexchadin: oh ok I see | 14:55 |
| openstackgerrit | Merged openstack/watcher-dashboard: Initializes EfficacyIndicator attributes https://review.openstack.org/339573 | 14:56 |
| alexchadin | vincentfrancoise: anyway, this way of implementation is more clean to me | 14:57 |
| vincentfrancoise | alexchadin: Something I would like to see in this spec is a decription/sequence diagram of how you intend to apply the JSON-like audit scope description to an in-memory model | 15:00 |
| vincentfrancoise | alexchadin: also, I believe that all cluster data models will have to respect a set of constraints in order for an audit scope to be applicable | 15:01 |
| *** acabot_ has quit IRC | 15:02 | |
| alexchadin | could you please give me more description about respecting of sets? | 15:03 |
| *** acabot_ has joined #openstack-watcher | 15:04 | |
| *** acabot_ has quit IRC | 15:05 | |
| vincentfrancoise | alexchadin: respecting of sets? | 15:05 |
| alexchadin | your last message :) | 15:05 |
| alexchadin | respecting constraints | 15:06 |
| vincentfrancoise | ok ok | 15:06 |
| vincentfrancoise | well, I guess you will need to traverse the model tree of elements to know whether or not you have to keep or filter out any given element. | 15:07 |
| vincentfrancoise | so that means you will need to know where to look within all these elements to find what you need to know isn't it? | 15:08 |
| vincentfrancoise | if so, we need to define the contract the CDMC developer will have to follow in order for the audit scope to work on custom models | 15:09 |
| vincentfrancoise | and that's this contract I'm referring to when I was talking about constraints | 15:09 |
| alexchadin | maybe we should create cluster model based on json (which already have been validated)? | 15:10 |
| vincentfrancoise | you mean, we take in in-memory model and only extract the bits we need given the audit scope? | 15:12 |
| alexchadin | yeap | 15:12 |
| vincentfrancoise | should work I guess | 15:12 |
| vincentfrancoise | then that's what I want to hear about in that spec in detail :) | 15:12 |
| alexchadin | vincentfrancoise: that's why I will review your BP this weekend :) | 15:13 |
| vincentfrancoise | that's the thing | 15:13 |
| alexchadin | cool | 15:13 |
| vincentfrancoise | I didn't detail the structure of the in-memory model | 15:14 |
| vincentfrancoise | and I was thinking about it this afternoon | 15:14 |
| vincentfrancoise | and so because nothing is yet set in stone, I think we should find a model structure that can fit both our BP needs | 15:15 |
| vincentfrancoise | here is an overview of the current model structure http://paste.openstack.org/show/528611/ | 15:16 |
| alexchadin | wow | 15:17 |
| alexchadin | great message! | 15:17 |
| vincentfrancoise | I want to think about something easy to maintain because right now it's too complicated | 15:17 |
| alexchadin | I agree | 15:17 |
| alexchadin | well | 15:17 |
| alexchadin | I will let you know my suggestions on Monday | 15:18 |
| alexchadin | will it suitable for you? | 15:18 |
| vincentfrancoise | but this is something we also need to talk about with jed56 which coming back wednesday | 15:19 |
| vincentfrancoise | yeah we can start discussing it together and expose something to him when he's back | 15:19 |
| alexchadin | would be good | 15:20 |
| vincentfrancoise | but keep in mind when designing the structure that we will have to explain how to implement one to a CDMC developer | 15:21 |
| vincentfrancoise | because right now the structure only fits to Nova | 15:21 |
| vincentfrancoise | we need something more generic | 15:21 |
| alexchadin | I got you, I've seen your updated docs already | 15:22 |
| vincentfrancoise | so there's many things to think about at the same time here... | 15:22 |
| alexchadin | well, I'm leaving my workplace | 15:24 |
| alexchadin | thank you for assistance and have a nice weekend! | 15:24 |
| openstackgerrit | David TARDIVEL proposed openstack/watcher-dashboard: Revert "Initializes EfficacyIndicator attributes" https://review.openstack.org/339618 | 15:28 |
| *** alexchadin has quit IRC | 15:29 | |
| openstackgerrit | Vladimir Ostroverkhov proposed openstack/watcher-dashboard: Add support continuously-optimization https://review.openstack.org/327095 | 15:37 |
| openstackgerrit | David TARDIVEL proposed openstack/watcher-dashboard: Revert "Initializes EfficacyIndicator attributes" https://review.openstack.org/339618 | 15:41 |
| *** vincentfrancoise has quit IRC | 15:54 | |
| sballe_ | Can somebody remind me what tox command I am suppose to run before submitting a patch? | 15:59 |
| openstackgerrit | Merged openstack/watcher-dashboard: Revert "Initializes EfficacyIndicator attributes" https://review.openstack.org/339618 | 15:59 |
| sballe_ | I mostly talking to check the style of the spec | 16:00 |
| dtardivel | sballe_: your patchset is related to which project ? | 16:00 |
| sballe_ | I am updating a spec in watcher-spec | 16:01 |
| dtardivel | you can just run 'tox' | 16:02 |
| sballe_ | so just tox and not tox -e py27 | 16:04 |
| dtardivel | tox will run all tox options | 16:07 |
| dtardivel | for spec we have tox -edocs I think | 16:08 |
| dtardivel | .... and tox -pep8 as well | 16:08 |
| sballe_ | ok thx | 16:09 |
| sballe_ | are you sre it is tox -pep8? I get command not found | 16:11 |
| ostrover_ | sballe_: maybe tox -e pep8 | 16:15 |
| openstackgerrit | Susanne Balle proposed openstack/watcher-specs: WiP Watcher should enforce the Nova policies https://review.openstack.org/329873 | 16:28 |
| *** ostrover_ has quit IRC | 16:52 | |
| diga | alexchadin : there ? | 16:52 |
| diga | acabot: what is the location where spec will get submitted ? | 17:00 |
| diga | acabot: https://github.com/openstack/watcher-specs/tree/master/specs/newton/approved | 17:06 |
| diga | correct ? | 17:06 |
| diga | acabot: got it | 17:09 |
| *** jinquan has quit IRC | 17:15 | |
| openstackgerrit | Susanne Balle proposed openstack/watcher-specs: WiP Watcher should enforce the Nova policies https://review.openstack.org/329873 | 18:06 |
| *** dtardivel has quit IRC | 18:27 | |
| openstackgerrit | Susanne Balle proposed openstack/watcher-specs: WiP Watcher should enforce the Nova policies https://review.openstack.org/329873 | 18:31 |
| *** esberglu has quit IRC | 18:56 | |
| *** esberglu has joined #openstack-watcher | 19:01 | |
| *** dronshaw has quit IRC | 19:11 | |
| *** wootehfoot has joined #openstack-watcher | 19:20 | |
| *** wootehfoot has quit IRC | 19:24 | |
| *** wootehfoot has joined #openstack-watcher | 19:27 | |
| *** jinquan1 has quit IRC | 19:42 | |
| *** jinquan has joined #openstack-watcher | 19:48 | |
| openstackgerrit | Daniel Pawlik proposed openstack/puppet-watcher: Fix authtoken options and adding the missing https://review.openstack.org/338585 | 20:23 |
| *** esberglu has quit IRC | 20:34 | |
| *** thorst has quit IRC | 20:40 | |
| *** esberglu has joined #openstack-watcher | 20:40 | |
| *** thorst has joined #openstack-watcher | 21:00 | |
| *** thorst has quit IRC | 21:05 | |
| *** wootehfoot has quit IRC | 21:05 | |
| *** esberglu has quit IRC | 21:22 | |
| openstackgerrit | Susanne Balle proposed openstack/watcher-specs: WiP Watcher should enforce the Nova policies https://review.openstack.org/329873 | 21:27 |
| openstackgerrit | Susanne Balle proposed openstack/watcher-specs: WiP Watcher should enforce the Nova policies https://review.openstack.org/329873 | 21:29 |
| openstackgerrit | Susanne Balle proposed openstack/watcher-specs: WiP Watcher should enforce the Nova policies https://review.openstack.org/329873 | 21:35 |
| *** figleaf is now known as edleafe | 22:16 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!