| *** thorst_away has joined #openstack-watcher | 00:24 | |
| *** thorst_away has quit IRC | 02:16 | |
| *** thorst_away has joined #openstack-watcher | 02:17 | |
| *** thorst_away has quit IRC | 02:25 | |
| *** esberglu has quit IRC | 02:49 | |
| *** thorst_away has joined #openstack-watcher | 03:23 | |
| *** thorst_away has quit IRC | 03:30 | |
| openstackgerrit | Tin Lam proposed openstack/watcher: Update docs links to docs.openstack.org https://review.openstack.org/333723 | 04:00 |
|---|---|---|
| *** thorst_away has joined #openstack-watcher | 04:28 | |
| *** thorst_away has quit IRC | 04:35 | |
| *** thorst_away has joined #openstack-watcher | 05:33 | |
| *** thorst_away has quit IRC | 05:40 | |
| *** thorst_away has joined #openstack-watcher | 06:38 | |
| *** thorst_away has quit IRC | 06:45 | |
| *** vtech has quit IRC | 06:45 | |
| *** pkoniszewski has quit IRC | 07:03 | |
| *** vtech has joined #openstack-watcher | 07:08 | |
| *** pkoniszewski has joined #openstack-watcher | 07:12 | |
| openstackgerrit | Merged openstack/python-watcherclient: Updated from global requirements https://review.openstack.org/332385 | 07:21 |
| openstackgerrit | Merged openstack/watcher: Updated from global requirements https://review.openstack.org/333711 | 07:21 |
| *** vincentfrancoise has joined #openstack-watcher | 07:27 | |
| *** alexchadin has joined #openstack-watcher | 07:37 | |
| *** vtech_ has joined #openstack-watcher | 07:39 | |
| *** vtech has quit IRC | 07:39 | |
| *** thorst_away has joined #openstack-watcher | 07:43 | |
| *** thorst_away has quit IRC | 07:50 | |
| openstackgerrit | Edwin Zhai proposed openstack/watcher: Enable strategy parameters https://review.openstack.org/297556 | 07:54 |
| vincentfrancoise | hi alexchadin | 08:42 |
| alexchadin | vincentfrancoise: hi | 08:42 |
| vincentfrancoise | I'm looking at the rally plugin but I can't run anything | 08:43 |
| alexchadin | Cause we haven't merged rally support for now | 08:44 |
| alexchadin | If you want, I can give you git patch:) | 08:45 |
| alexchadin | Once you apply it you will able to run watcher scenarios in rally | 08:45 |
| vincentfrancoise | I checked out your review to run it | 08:46 |
| vincentfrancoise | Can you push the missing code for review then ? | 08:46 |
| alexchadin | I need to push that code to rally repo | 08:47 |
| alexchadin | To do this I need to set rally gate up | 08:47 |
| vincentfrancoise | I thought it was done like in http://docs.openstack.org/developer/rally/plugins/scenario_plugin.html | 08:47 |
| vincentfrancoise | but with the code being hosted in the watcher codebase | 08:48 |
| vincentfrancoise | I am using docker rally to run the tests | 08:48 |
| alexchadin | They prefer to host code in their repo | 08:48 |
| *** thorst_away has joined #openstack-watcher | 08:48 | |
| alexchadin | to make it "upstream" | 08:48 |
| vincentfrancoise | Oh I see | 08:48 |
| vincentfrancoise | Put me as reviewer when you push for review that changeset then | 08:49 |
| alexchadin | vincentfrancoise: ok :) | 08:50 |
| vincentfrancoise | also, can you also put the command lines into the README.rst so we can run rally tests locally? | 08:50 |
| alexchadin | what sort of commands? | 08:51 |
| vincentfrancoise | a simplified version of what is in http://docs.openstack.org/developer/rally/install.html | 08:52 |
| vincentfrancoise | and probably a link to this page as well | 08:52 |
| alexchadin | Oh | 08:52 |
| alexchadin | Of course | 08:53 |
| vincentfrancoise | but the the command lines tuned to run the Watcher tests | 08:53 |
| alexchadin | I understand | 08:53 |
| alexchadin | while you are here | 08:53 |
| alexchadin | I have some questions:) | 08:53 |
| vincentfrancoise | what is it ? | 08:53 |
| alexchadin | regarding cont-opt | 08:53 |
| vincentfrancoise | ah yeah writing the comments on this was not easy so I was probably not explicit/clear enough | 08:54 |
| alexchadin | Should AuditHandler class be metaclass? | 08:54 |
| vincentfrancoise | I think so | 08:55 |
| *** thorst_away has quit IRC | 08:55 | |
| alexchadin | okay | 08:55 |
| alexchadin | then | 08:55 |
| vincentfrancoise | AuditHandler is just to avoid too much code duplication | 08:55 |
| alexchadin | the execite logic | 08:55 |
| vincentfrancoise | like BaseAuditHandler would be a Java Interface | 08:55 |
| alexchadin | I see there is 2 way to get it refactored | 08:56 |
| vincentfrancoise | whilst AuditHandler is the actual abstract class | 08:56 |
| alexchadin | hm | 08:56 |
| alexchadin | The difference between metaclass and abstract class | 08:56 |
| alexchadin | metaclass has logic in its methods | 08:57 |
| alexchadin | abstract class is something like interfaces in Python | 08:57 |
| vincentfrancoise | Oh I thought you were referring to the same thing since you make an abstract class via the six.add_metaclass() decorator | 08:58 |
| alexchadin | So we are using six.add_metaclass on AuditHandler, right? | 08:59 |
| vincentfrancoise | yes | 09:00 |
| alexchadin | cool | 09:00 |
| vincentfrancoise | BTW what do you think about doing it this way? | 09:00 |
| alexchadin | I like your idea with inheritance | 09:01 |
| alexchadin | BTW I have questons about execute method | 09:01 |
| vincentfrancoise | ^ cool :) | 09:02 |
| alexchadin | In both ways (oneshot, continuous) we have to update audit state, call planner to schedule, etc | 09:03 |
| vincentfrancoise | ah yes | 09:03 |
| alexchadin | There could be duplicates in execute method | 09:03 |
| alexchadin | :) | 09:03 |
| vincentfrancoise | I was thinking about this when writing my comments | 09:03 |
| vincentfrancoise | I was thinking we could do some non abstract execute() method that call pre_execute() do_execute() and post_execute() | 09:04 |
| vincentfrancoise | duplicate logic would then go into the pre and post with only the do_execute() differing | 09:05 |
| alexchadin | hm | 09:05 |
| vincentfrancoise | pre_execute() do_execute() and post_execute() would then be abstract in BaseAuditHandler | 09:05 |
| vincentfrancoise | and pre_execute() and post_execute() could be implemented in AuditHandler | 09:06 |
| vincentfrancoise | what do you think? | 09:06 |
| alexchadin | then do_execute would be empty for oneshot method since there is only one difference between them | 09:06 |
| alexchadin | if statement for continuous | 09:07 |
| vincentfrancoise | oneshot would only do self.strategy_context.execute_strategy() | 09:07 |
| vincentfrancoise | while continuous would also self.strategy_context.execute_strategy() but also the extra bit | 09:07 |
| alexchadin | yes | 09:10 |
| alexchadin | That's definitely what I wanted | 09:11 |
| alexchadin | You confirmed my thoughts | 09:11 |
| *** hvprash has joined #openstack-watcher | 09:12 | |
| *** hvprash has quit IRC | 09:17 | |
| vincentfrancoise | cool :) | 09:18 |
| *** thorst_away has joined #openstack-watcher | 09:52 | |
| *** thorst_away has quit IRC | 10:00 | |
| *** alexchadin has quit IRC | 10:14 | |
| *** alexchadin has joined #openstack-watcher | 10:14 | |
| *** vincentfrancoise has quit IRC | 10:16 | |
| alexchadin | jed56: are you here? | 10:37 |
| *** wootehfoot has joined #openstack-watcher | 10:51 | |
| *** acabot has quit IRC | 10:54 | |
| *** thorst_away has joined #openstack-watcher | 10:58 | |
| *** thorst_away has quit IRC | 11:05 | |
| *** acabot has joined #openstack-watcher | 11:06 | |
| *** thorst_away has joined #openstack-watcher | 11:07 | |
| *** thorst_away is now known as thorst_ | 11:15 | |
| *** hvprash has joined #openstack-watcher | 11:41 | |
| *** hvprash has quit IRC | 11:47 | |
| *** hvprash has joined #openstack-watcher | 11:47 | |
| *** hvprash_ has joined #openstack-watcher | 11:48 | |
| *** hvprash has quit IRC | 11:52 | |
| *** hvprash_ has quit IRC | 12:00 | |
| *** hvprash has joined #openstack-watcher | 12:00 | |
| jed56 | alexchadin: hello | 12:05 |
| jed56 | I'm here :) | 12:05 |
| alexchadin | jed56: hi:) | 12:06 |
| alexchadin | I think about layer where we should have our audit handlers | 12:07 |
| jed56 | what do you mean ? | 12:08 |
| alexchadin | For now we have auditendpoint which is RPC enpoint | 12:08 |
| alexchadin | I agree it is wrong way to store default.ContinuousAuditManager in endpoints | 12:08 |
| alexchadin | https://review.openstack.org/#/c/321411/6/watcher/decision_engine/manager.py | 12:08 |
| jed56 | yes we discussed that we vincentfrancois | 12:09 |
| jed56 | okay | 12:10 |
| alexchadin | Maybe we should have something like AuditManager | 12:12 |
| alexchadin | Where we can dispatch audit to the right handler | 12:12 |
| jed56 | You don't want to put the switch logic in the AuditEndpoint ? | 12:14 |
| jed56 | We can imagine to create an AuditManager if we have maybe more than two types of audits | 12:15 |
| *** dtardivel has joined #openstack-watcher | 12:15 | |
| jed56 | https://github.com/openstack/watcher/blob/91951f3b012c10c21da2cde37cffbeba0208fc84/watcher/decision_engine/messaging/audit_endpoint.py#L45 | 12:15 |
| jed56 | It is sad that we can't easily implement inheritance with the openstack objects | 12:17 |
| jed56 | We could create a OneshotAudit and ContinuousAudit and use polymorphism to decided what is the handler to use | 12:17 |
| alexchadin | jed56: If we put switch logic in the AuditEndpoint we need to know audit type | 12:18 |
| *** vincentfrancoise has joined #openstack-watcher | 12:18 | |
| jed56 | why ? | 12:19 |
| jed56 | You query the db to have this information | 12:19 |
| jed56 | *can | 12:19 |
| alexchadin | yes | 12:19 |
| alexchadin | hm | 12:19 |
| jed56 | There is a reason why you don't want to do that ? | 12:20 |
| alexchadin | There isn't, but I always tried to reduce number of queries to the DB | 12:21 |
| alexchadin | To make code a little faster | 12:21 |
| jed56 | alexchadin: yes this is good idea to reduce the number of queries | 12:21 |
| jed56 | let's me take a look to the code | 12:22 |
| jed56 | IMHO, we can move the query in order to do it only once | 12:22 |
| alexchadin | We can send audit object | 12:22 |
| alexchadin | to the execute function | 12:22 |
| jed56 | so, the query is made by the DefaultStrategyContext class | 12:23 |
| *** hvprash has quit IRC | 12:23 | |
| *** hvprash has joined #openstack-watcher | 12:23 | |
| jed56 | We can move this logic to the AuditManager | 12:24 |
| jed56 | and like you said pass the object and not anymore the audit_uuid | 12:24 |
| alexchadin | jed56: yeap:) | 12:24 |
| jed56 | I don't think this will cause problems | 12:25 |
| alexchadin | thank you for your help | 12:25 |
| jed56 | welcome ! | 12:25 |
| vincentfrancoise | alexchadin: I juste tried to run your rally tests again but I get errors | 12:35 |
| *** hvprash has quit IRC | 12:35 | |
| alexchadin | which types of errors? | 12:35 |
| vincentfrancoise | http://pastebin.com/utpHRgd6 | 12:35 |
| alexchadin | thank you for bug | 12:37 |
| alexchadin | s/audit_templates/audit_template | 12:37 |
| alexchadin | I forgot to change it in basic.py | 12:37 |
| alexchadin | vincentfrancoise: have you changed it? | 12:39 |
| *** alexchadin has quit IRC | 12:42 | |
| *** hvprash has joined #openstack-watcher | 12:57 | |
| vincentfrancoise | alexchadin: http://pastebin.com/aGAaSCqx | 13:00 |
| *** edleafe is now known as figleaf | 13:01 | |
| *** hvprash has quit IRC | 13:02 | |
| *** esberglu has joined #openstack-watcher | 13:19 | |
| *** hvprash has joined #openstack-watcher | 13:20 | |
| *** vtech_ has quit IRC | 13:23 | |
| *** hvprash_ has joined #openstack-watcher | 13:24 | |
| *** esberglu has quit IRC | 13:24 | |
| *** hvprash has quit IRC | 13:24 | |
| *** hvprash has joined #openstack-watcher | 13:27 | |
| *** dronshaw has joined #openstack-watcher | 13:30 | |
| *** hvprash_ has quit IRC | 13:30 | |
| figleaf | jed56: I've added your spec to the agenda for next Monday's scheduler meeting | 13:34 |
| figleaf | jed56: https://wiki.openstack.org/wiki/Meetings/NovaScheduler#Agenda_for_next_meeting | 13:35 |
| figleaf | jed56: I want to get a clear answer from the scheduler subteam on whether your approach is something that we should encourage or discourage | 13:35 |
| figleaf | jed56: It's at 1400 UTC in #openstack-meeting-alt | 13:36 |
| jed56 | figleaf: thanks you very much | 13:36 |
| jed56 | we will be there with acabot | 13:36 |
| figleaf | jed56: great! | 13:36 |
| jed56 | thank you very much for that | 13:37 |
| *** esberglu has joined #openstack-watcher | 13:39 | |
| jed56 | jinquan: hi | 13:39 |
| jed56 | jinquan: do you feel confident to develop the configurable-weights-default-planner blueprint without specification ? | 13:40 |
| jed56 | jinquan: IMHO, we didn't need a spec but if you need one I can write one. | 13:41 |
| jed56 | jinquan: if you have any questions don't hesitate | 13:41 |
| *** vtech has joined #openstack-watcher | 13:46 | |
| *** vtech has quit IRC | 13:55 | |
| *** ecelik has quit IRC | 13:57 | |
| *** jwcroppe has joined #openstack-watcher | 14:05 | |
| *** vtech has joined #openstack-watcher | 14:12 | |
| *** brunograz has quit IRC | 15:30 | |
| *** vtech has quit IRC | 15:33 | |
| *** vtech has joined #openstack-watcher | 15:54 | |
| *** wootehfoot has quit IRC | 15:57 | |
| *** vtech has quit IRC | 16:17 | |
| openstackgerrit | Vincent Françoise proposed openstack/watcher: Loadable Cluster Data Model Collectors https://review.openstack.org/333987 | 16:27 |
| *** vincentfrancoise has quit IRC | 16:30 | |
| *** acabot has quit IRC | 16:50 | |
| *** acabot has joined #openstack-watcher | 17:01 | |
| *** jed56 has quit IRC | 17:15 | |
| *** dtardivel has quit IRC | 17:17 | |
| *** vtech has joined #openstack-watcher | 17:41 | |
| *** openstackgerrit has quit IRC | 18:03 | |
| *** openstackgerrit has joined #openstack-watcher | 18:03 | |
| *** vtech has quit IRC | 18:22 | |
| *** openstack has joined #openstack-watcher | 18:41 | |
| *** dronshaw has quit IRC | 19:21 | |
| *** esberglu has quit IRC | 19:26 | |
| *** esberglu has joined #openstack-watcher | 19:34 | |
| *** esberglu has quit IRC | 21:07 | |
| *** esberglu has joined #openstack-watcher | 21:14 | |
| *** esberglu has quit IRC | 21:14 | |
| *** thorst_ has quit IRC | 21:49 | |
| *** thorst has joined #openstack-watcher | 21:50 | |
| *** figleaf is now known as edleafe | 21:52 | |
| *** thorst has quit IRC | 21:58 | |
| *** harlowja has quit IRC | 22:17 | |
| *** harlowja has joined #openstack-watcher | 22:30 | |
| *** thorst has joined #openstack-watcher | 22:57 | |
| *** thorst has quit IRC | 23:04 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!