08:00:51 #startmeeting vitrage 08:00:52 Meeting started Wed Aug 1 08:00:51 2018 UTC and is due to finish in 60 minutes. The chair is ifat_afek. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:00:54 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 08:00:56 The meeting name has been set to 'vitrage' 08:00:59 Hi :-) 08:01:05 hi 08:01:10 <_Adary> mornin' 08:01:24 hi 08:01:36 hi 08:01:48 \o/ 08:02:02 hello hello 08:02:22 <_Adary> °º¤ø,¸¸,ø¤º°`°º¤ø,¸,ø¤°º¤ø,¸¸,ø¤º°`°º¤ø,¸ 08:02:30 cool!! 08:02:54 A warning before we start: there is a spam attack on our IRC channels today. I hope it won’t bother us too much 08:03:20 Agenda: 08:03:26 * Status and updates 08:03:32 * Vitrage virtual PTG 08:03:36 * Open discussion 08:03:44 #topic Status and updates 08:03:51 Last week we released python-vitrageclient 2.3.0 08:03:59 The API is final, but the related changes in vitrage and vitrage-dashboard are still missing. We hope to release them in a day or two. 08:04:28 Next milestones: 08:04:32 Release candidate on August 9 08:04:36 Final release on August 23 08:04:40 #link https://releases.openstack.org/rocky/schedule.html 08:04:45 #link https://storyboard.openstack.org/#!/board/90 08:04:56 My updates: 08:05:03 I’m working on a solution to the datasources cache problem that was created as a result of persisting Virage state in the DB. 08:05:15 Many datasources have caches, and they are using them (among other things) to determine that an entity was deleted. There are two problems with it: 08:05:21 1. If there are multiple vitrage-collector processes (HA mode), not all caches will be updated 08:05:27 1. If an entity is deleted when the vitrage-graph is down, the cache will be empty so the datasource will not know that the entity was deleted 08:05:42 The solution for Rocky: store on each entity the name of the datasource. Add a should_delete_outdated_entities() method that returns false by default. And let the consistency service delete those entities (after 10 minutes or so). 08:05:49 The solution for Stein: The processor should delete the outdated entities immediately after get_all. 08:05:57 Other than that, I’m fixing code review issues in the Vitrage resource in Heat. 08:06:01 That’s it for me. 08:06:34 I'll update 08:07:35 I'm still working on alarm history, working on last few things, hoping to push it by tomorrow for review 08:07:57 do we need feature freeze exception for it? 08:08:25 the changes in UI are already ready for the history feature, they will be pushed after I'll push my feature 08:08:48 e0ne: I’m not sure if we officially need it, since we are working on a cycle-with-milestones mode. But it is a new feature, yes 08:09:02 I meand cycle-with-intermediary 08:10:12 ifat_afek: ok, thanks for the information 08:10:42 e0ne: we have another feature that is not done yet, the Prometheus datasource. I’ll check regarding the feature freeze extenstion policy 08:11:30 tanks ifat_afek :) 08:11:36 thats it for me 08:12:16 My update 08:12:24 The fast failover commits are merged 08:12:30 #link https://review.openstack.org/#/q/status:merged+project:openstack/vitrage+branch:master+topic:idandos/init_from_db 08:12:38 I have a few more minor issues to attend. 08:12:41 Great :-) 08:12:51 +2 08:13:03 ifat_afek: if we use cycle-with-intermediary, we don't need FFE 08:13:21 e0ne: thanks for the info 08:13:25 2np 08:13:27 np 08:13:28 Also, 08:13:37 If you pull the recent code from master to your devstack, make sure to run vitrage-dbsync to upgrade your db. 08:13:48 Mostly for development reasons, there is a new script vitrage-purge-data, make sure to run it after stopping all vitrage services. 08:13:55 To get this script, run 'sudo pip install -e .' in vitrage folder. 08:14:15 idan_hefetz: Thanks. I think we should send this info in the mailing list 08:14:37 OK, cool 08:14:56 That's it for me 08:15:14 Hey 08:15:27 Im adding buttons to vitrage dashboard to add and delete templates 08:15:30 should be done today 08:15:40 and thats it for me :) 08:15:50 Cool 08:16:14 * e0ne is always happy to get new UI features landed 08:16:40 :-) 08:17:11 Unfortunately we don’t have enough UI resources 08:17:19 We could do a lot more 08:17:38 :( 08:18:07 An update on behalf of najjar: I know that he finished the work on Prometheus datasource (notifications only). He is in code review phase. Then he plans to support get_all for Prometheus 08:18:58 Anyone else wants to update? 08:20:01 he pushed the basic auth middleware 08:20:32 eyalb: right. And he has another change waiting to be pushed, the last one for this feature 08:21:22 since prometeus uses basic auth for authentication we need to support it for vitrage 08:21:31 #topic Vitrage virtual PTG 08:21:49 We already discussed the issue that Vitrage will hold a virtual PTG instead of participating in the Denver PTG 08:21:59 Regarding the date, due to several restrictions I thought of holding it on the first week of October. Does that work for everyone? 08:22:20 Sorry I'm late. 08:22:30 MinWookKim: hi 08:22:51 ifat_afek: Hello : ) 08:22:58 I’ll repeat my last question: we are thinking of holding a Vitrage virtual PTG on the first week of October. Does that work for you? 08:23:14 +1 08:23:19 <_Adary> +2 08:23:22 I’ll send a doodle poll regarding the specific dates later today 08:23:26 +1 08:23:32 And an important issue is the time zones… what are yours? 08:24:01 We are currently UTC+3, might be changed to UTC+2 in October (not sure when the daylight saving time ends) 08:24:10 ifat_afek: UTC+3, I think, it's the same as yours 08:25:07 I created an etherpad where you can start adding your ideas for discussions: 08:25:11 #link https://etherpad.openstack.org/p/vitrage-stein-ptg 08:25:24 I only listed one idea for now. It is specifically relevant to my development in Heat 08:26:18 #topic Open discussion 08:26:23 Any other issues? 08:26:31 I've got a question about this commit: https://review.openstack.org/#/c/586199/3/vitrage/datasources/static_physical/__init__.py 08:26:46 ifat_afek, eyalb: it confused me a lot:( 08:27:13 this is something that is not used anymore and should be deleted 08:27:32 patch is merged in Rocky, but deprecation message says it's deprecated in Queens or Pike 08:28:39 eyalb: I'm all for removing dead code, but it would be good to have a clear message since what release it has been deprecated 08:28:56 e0ne: The purpose of the static datasource is to add resources that are not changed often and are hard to auto-discover, like switches and other network elements. Practically, we use it mostly for demos. The static_physical is an old version of the datasource with many limitations. The static datasource replaced it in Pike 08:29:28 eOne: I will change it to queens 08:30:07 this line https://github.com/openstack/vitrage/blob/master/vitrage/datasources/static_physical/__init__.py#L27 doesn't fit to https://github.com/openstack/vitrage/blob/master/vitrage/datasources/static_physical/__init__.py#L38 08:30:17 that's why I raised this question 08:30:29 eyalb: thanks! 08:31:08 With our IRC ad service you can reach a global audience of entrepreneurs and fentanyl addicts with extraordinary engagement rates! https://williampitcock.com/ 08:31:11 I thought you guys might be interested in this blog by freenode staff member Bryan 'kloeri' Ostergaard https://bryanostergaard.com/ 08:31:15 Read what IRC investigative journalists have uncovered on the freenode pedophilia scandal https://encyclopediadramatica.rs/Freenodegate 08:31:15 And here comes the spam :-( 08:31:18 A fascinating blog by freenode staff member Matthew 'mst' Trout https://MattSTrout.com/ 08:31:31 Please ignore... 08:31:36 Any other open issues? 08:31:43 nothing from me 08:33:15 I guess we are done, have a nice day :-) 08:33:25 bye 08:33:27 <_Adary> buhbye 08:33:38 See you : ) 08:33:49 #endmeeting