Wednesday, 2016-06-15

*** thorst has joined #openstack-watcher00:06
*** thorst has quit IRC00:10
*** thorst has joined #openstack-watcher00:53
*** hvprash has joined #openstack-watcher00:59
*** hvprash has quit IRC01:22
*** hvprash has joined #openstack-watcher01:23
*** hvprash has quit IRC01:26
*** hvprash has joined #openstack-watcher01:27
*** thorst has quit IRC01:36
*** thorst has joined #openstack-watcher01:37
*** hvprash has quit IRC01:42
*** M00nr41n has joined #openstack-watcher01:43
*** thorst has quit IRC01:46
*** hvprash has joined #openstack-watcher02:02
*** hvprash has quit IRC02:07
*** thorst has joined #openstack-watcher02:28
*** thorst has quit IRC02:28
*** thorst has joined #openstack-watcher02:29
*** thorst has quit IRC02:37
*** thorst has joined #openstack-watcher03:35
*** thorst has quit IRC03:43
*** M00nr41n has quit IRC03:43
*** jwcroppe has joined #openstack-watcher04:17
*** thorst has joined #openstack-watcher04:42
*** thorst has quit IRC04:48
*** M00nr41n has joined #openstack-watcher05:30
*** M00nr41n has quit IRC05:42
*** thorst has joined #openstack-watcher05:47
*** thorst has quit IRC05:53
*** openstackgerrit has quit IRC06:18
*** openstackgerrit has joined #openstack-watcher06:18
*** jinquan1 has joined #openstack-watcher06:18
*** vtech has quit IRC06:44
*** thorst has joined #openstack-watcher06:51
*** thorst has quit IRC06:58
*** vtech has joined #openstack-watcher07:05
*** dtardivel has joined #openstack-watcher07:14
*** danpawlik has joined #openstack-watcher07:25
*** openstackgerrit has quit IRC07:33
*** openstackgerrit has joined #openstack-watcher07:33
*** alexchadin has joined #openstack-watcher07:38
*** AJaeger has joined #openstack-watcher07:39
alexchadinjed56: how are you?07:39
jed56I'm doing well and you ?07:39
danpawlikhello guys07:39
jed56hello danpawlik07:39
alexchadinjed56: thanks you for your review07:39
alexchadindanpawlik: hi07:40
alexchadinjed56: I'm fine07:40
jed56alexchadin: I'm currently testing the apscheduler library07:40
openstackgerritAndreas Jaeger proposed openstack/watcher: Fix releasenotes generation
alexchadinjed56: what can you say about it?07:41
AJaegeracabot: tox -e releasenotes fails completely, I've fixed it for you ^07:41
AJaegeracabot: without working releasenotes, the translation sync fails as well07:41
acabotAJaeger : ok let met see that07:42
jed56alexchadin:  it's seem that we can use
jed56the same approach for schedule the audits07:42
jed56we have to figure out what is the best scheduler for our use case07:44
jed56tornado, gevent, etc07:45
jed56do you want to take a look ?07:45
alexchadinjed56: I want to release support for rally asap, some unit tests left07:47
alexchadinjed56: why can't we just take apscheduler?07:47
AJaegeracabot: if you have questions about the change, just ask...07:49
*** jwcroppe has quit IRC07:51
*** M00nr41n has joined #openstack-watcher07:53
*** thorst has joined #openstack-watcher07:55
alexchadinjed56: you mean support of apscheduler internal classes?07:58
alexchadinjed56: like tornadoscheduler, geventscheduler07:59
*** thorst has quit IRC08:02
jed56alexchadin: yes08:05
jed56we should carefully select the best scheduler08:05
alexchadinjed56: do we use any of these frameworks?08:07
jed56In watcher ?08:08
jed56no, I pretty sure that we are not using these frameworks08:09
jed56I will test today these frameworks and i will let you the best one08:11
alexchadinjed56: doc of apscheduler says that we should use appropriate scheduler08:11
alexchadinsorry for my misunderstanding:)08:12
alexchadinbut it seem that BackgroundScheduler is right choice for us, no?08:12
jed56coffee break :)08:12
alexchadinregarding our stress discussion: take a look:
openstackgerritAndreas Jaeger proposed openstack/watcher: Use proper theme for release notes
openstackgerritEdwin Zhai proposed openstack/watcher: Enable strategy parameters
jed56alexchadin: we know the stress tool08:28
jed56IMO, we will use
jed56The scope of the stress is only CPU08:29
*** alexchadin has quit IRC08:34
openstackgerritMerged openstack/watcher: Fix releasenotes generation
openstackgerritDaniel Pawlik proposed openstack/puppet-watcher: Implement api.pp
AJaegerwatcher releasenotes are published now for the first time at
AJaegerNow you only need to write some ;)08:56
jed56AJaeger: great! Thanks a lot :)08:57
acabotAJaeger : thx !08:58
AJaegeryou're welcome08:59
openstackgerritMerged openstack/python-watcherclient: Updated CLI to display efficacy related fields
AJaegerhave a great day...08:59
*** AJaeger has left #openstack-watcher08:59
openstackgerritMerged openstack/watcher: Decoupled Goal from Strategy
openstackgerritMerged openstack/watcher: Added EfficacyIndicator model in DB
jed56you too08:59
openstackgerritMerged openstack/watcher: Added efficacy specification to /goals
openstackgerritMerged openstack/watcher: Added EfficacyIndicator object
openstackgerritMerged openstack/watcher: Added pre/post execution methods to strategies
openstackgerritMerged openstack/watcher: Added efficacy indicators to /action_plans
openstackgerritMerged openstack/watcher: Documentation on goal and efficacy
openstackgerritMerged openstack/watcher-dashboard: Added efficacy-related fields to the dashboard
*** thorst has joined #openstack-watcher09:00
openstackgerritMerged openstack/watcher: Use proper theme for release notes
*** thorst has quit IRC09:07
*** ostroverkhov has joined #openstack-watcher09:27
*** ostroverkhov has quit IRC09:28
brunograzHi all09:38
brunograzI just noticed that some of my review code could not be merged09:39
brunograzis there a way to solve this?09:40
brunograzI also created 2 more bugs regarding our strategy - and
openstackLaunchpad bug 1591901 in watcher "vm workload strategy does not change nova service state" [High,New] - Assigned to Jinquan Ni (ni-jinquan)09:41
openstackLaunchpad bug 1591927 in watcher "Vm workload consolidation strategy tries to disable already disabled nova-compute services" [Wishlist,New]09:41
brunograzI'm just waiting until it gets triaged so I can strat working on them09:41
jed56hello brunograz09:44
jed56you should rebase your patchset09:44
brunograzhi jed5609:45
brunograzis there a specific PS to use?09:45
jed56what do you mean ? :)09:46
brunograzto rebase the code09:46
brunograzI cannot rebase due merge conflicts09:48
jed56You have to resolve the conflicts after you git rebase09:51
jed56Git is telling you which files are causing the conflicts09:52
brunograzI guess it should be discussed in the meeting today10:01
brunograzbecause the latest version of the watcher-dashboard in the git repo has been updated 5 days ago and it implements the bug reported10:02
brunograzmaybe we should change it status to 'Invalid'10:02
jed56brunograz: ah, this is maybe vincentfrancoise  who fixed the problem.10:04
jed56sorry about that10:05
*** thorst has joined #openstack-watcher10:06
brunograzalso for the other bug report10:06
jed56I agree we should set the status to Invalid10:06
brunograznot a problem ;)10:06
brunograzjust that it was not clear for me that there were ppl working in the same issue10:07
brunograzactually merged few minutes ago
*** thorst has quit IRC10:12
acabotagenda for today #link
acabotI probably need to update the watcher-dashboard section10:13
acabotbrunograz : can you do some clean-up on your patches ?10:14
brunograzacabot: I marked the bug as invalid10:14
brunograzwill check how to remove the PS10:14
acabotbrunograz : ok you can simply abandon your patch10:14
acabotsorry about that10:14
brunograzacabot: done ;)10:15
acabotgzhai2 : I'm really sorry about that but you will have to rebase again10:20
*** jwcroppe has joined #openstack-watcher10:31
jinquan1brunograz: hello?10:31
brunograzjinquan1: hi10:31
jinquan1i read the chat log, are you mean that to work the bug
openstackLaunchpad bug 1591901 in watcher "vm workload strategy does not change nova service state" [High,Triaged] - Assigned to Jinquan Ni (ni-jinquan)10:32
jinquan1if yes  i will unssign10:32
brunograzjinquan1: if you would like to work on that bug for me isn't an issue10:33
brunograzwe implemented that strategy so I guess it would be easier for us to find the bug10:33
brunograzbut if you want to work on it go for it10:34
jinquan1 brunograz: ok10:37
jinquan1get it :)10:37
openstackgerritJean-Emile DARTOIS proposed openstack/watcher-specs: WiP Watcher should respect the Nova policies
*** hvprash has joined #openstack-watcher10:50
*** hvprash has quit IRC10:54
openstackgerritJean-Emile DARTOIS proposed openstack/watcher: Fix StrategyContext to use the strategy_id in the Audit Template
*** thorst has joined #openstack-watcher11:10
danpawlikSpecs file for openstack-watcher created :D Today I try to push it to community ;)11:10
*** thorst has quit IRC11:17
*** alexchadin has joined #openstack-watcher11:18
jed56danpawlik: great11:41
*** thorst has joined #openstack-watcher11:47
alexchadinjed56: should we store our scheduler jobs in somewhere?12:05
alexchadinsql db, mongo12:05
jed56alexchadin: I'm currently amending the specs :)12:05
jed56I'm claim that for the first implementation we should use in memory12:06
alexchadinjed56: good:)12:07
alexchadinWhat did you mean by "You don't take into account that each audit can have different periods."12:07
alexchadinEach audit have only one period, no?12:08
*** jwcroppe has quit IRC12:10
jed56Each audit have only one period, no? yes, but in your implementation you only relies on CONF.period_interval.12:11
jed56you have to use for each audit his own period12:12
alexchadinI use12:13
alexchadinWe had discussion with vincentfrancoise about this12:14
jed56ah yes, i didn't saw this line sorryu12:14
alexchadinI will think of it BTW12:14
*** alexchadin has quit IRC12:28
*** Zucan has joined #openstack-watcher12:34
*** M00nr41n has left #openstack-watcher12:36
*** hvprash has joined #openstack-watcher13:00
*** hvprash has quit IRC13:01
*** jwcroppe has joined #openstack-watcher13:11
jed56sballe_: hi13:22
sballe_I am going to setup Watcher and devstack. I am assuming the documentation is on our normal wiki13:23
sballe_Any best practices around how to set that up given that devstack can be setup with more than one VM13:23
*** michaelgugino has joined #openstack-watcher13:24
jed56Personally, I use a vagrant script13:37
sballe_Is the vagrant approach documented somewhere?13:39
tkaczynskisballe_ I hope that things have improved, I just remember that I spent like 3 weeks trying to setup multinode devstack not that long ago13:52
*** hvprash has joined #openstack-watcher13:55
*** vtech has quit IRC13:55
sballe_me 213:56
*** hvprash has quit IRC13:58
*** hvprash has joined #openstack-watcher13:58
acabotweekly meeting will start on #openstack-meeting-414:00
*** brunograz has quit IRC14:00
*** jed56 has quit IRC14:01
*** jed56 has joined #openstack-watcher14:01
*** brunograz has joined #openstack-watcher14:01
*** vtech has joined #openstack-watcher14:02
*** ecelik has quit IRC14:02
*** vtech has quit IRC14:04
openstackgerritOpenStack Proposal Bot proposed openstack/watcher: Updated from global requirements
*** vtech has joined #openstack-watcher14:05
*** vtech has quit IRC14:05
*** vtech has joined #openstack-watcher14:06
*** vtech has quit IRC14:07
*** vtech has joined #openstack-watcher14:17
*** vtech has quit IRC14:20
*** vtech has joined #openstack-watcher14:23
*** vtech has quit IRC14:25
*** vtech has joined #openstack-watcher14:28
*** jwcroppe has quit IRC14:52
*** vtech has quit IRC14:53
*** vtech has joined #openstack-watcher14:55
*** brunograz has quit IRC14:58
*** vtech has quit IRC14:59
*** odyssey4me has joined #openstack-watcher15:00
*** brunograz has joined #openstack-watcher15:00
odyssey4mehvprash acabot I saw some discussion in #openstack-meeting-4 with regards to getting Watcher deployable with OSA.15:01
hvprashwould like to talk to you about that15:02
odyssey4meThis has been mentioned by michaelgugino who's plugged into the OSA development community, and I see some others here who've expressed interest in Watcher being deployed in OSA.15:02
odyssey4meSure, interest has definitely been expressed. How far along is watcher?15:03
acabotodyssey4me : watcher is a big tent project for 2 weeks now15:04
odyssey4meThe ideal situation is to have someone who understands the service very well to work with someone who understand OSA very well in a kind-of partnership so that any changes to Watcher can quickly be implemented in OSA rather than being reactive.15:04
acabotodyssey4me : sure I understand15:05
odyssey4meacabot sure, but what I mean is how well documented is the deployment, how usable is it for production?15:05
acabotodyssey4me : so the best contact for this would be dtardivel who has experience with ansible15:05
michaelguginoodyssey4me: that will be me, I'm going to be submitting patches against watcher with hvprash15:05
odyssey4memy apologies, I haven't gone and looked through any published info about it15:05
michaelguginoor dtardivel can do it ;)15:05
acabotmichaelgugino : no I mean dtardivel understand the service and should work with you as you know OSA15:06
odyssey4mesounds good to me15:06
acabotodyssey4me :
acabotodyssey4me : watcher can be deployed with puppet or in a devstack env15:07
acabotodyssey4me : its not production ready as it is a 0.27 version right now15:07
odyssey4meacabot sure, that's not an issue - we can note it as an 'experimental' feature15:08
acabotodyssey4me : it would make sense15:08
acabotsame state on puppet side ;-)15:08
odyssey4meyeah, Emilien and I have been trading notes now and again :)15:09
acabotok ;-)15:09
odyssey4meok, I guess it's really going to come down to who has time to do what, when15:09
acabotmichaelgugino : when do you plan to work on it ?15:09
odyssey4memichaelgugino I'll leave it to you to work with the Watcher dev team to make it happen.15:09
hvprashwe have few more members who can work on this15:10
odyssey4meacabot Feel free to ping me if the Watcher team needs a fresh set of eyes or if things are a bit stuck.15:10
michaelguginoacabot: I'm planning on helping hvprash with the audit modification work this week.  I'll be on holiday next week, so most likely the week after that I'll start putting together some plays for osa-watcher15:10
hvprashthx michaelgugino odyssey4me15:11
acabotmichaelgugino : ok in this case I will ask dtardivel to look at it in 2 weeks15:11
michaelguginook, sounds good15:11
odyssey4melooking forward to it :)15:12
odyssey4meon paper it looks interesting, so I look forward to seeing it running15:15
openstackgerritDaniel Pawlik proposed openstack/puppet-watcher: Implement api.pp
*** danpawlik has quit IRC15:20
edleafereading the scrollback from the meeting - is there still a question as to what a "fishbowl session" is at the summit?15:31
jed56edleafe: +1 :)15:32
jed56acabot: can you explain to us what is a fishbolw session ?15:32
edleafeA fishbowl session is one where there is a person or two leading the discussion, and everyone else arranged in a semi-circular ring around them15:33
edleafeThose who are most involved with the topic sit close; those who are lurking sit further away15:34
edleafeYou can move, of course, if you realize that you either want to contribute more (or less)15:34
edleafeThere are no microphones; people are supposed to sit close enough so that everyone can hear15:35
edleafeIt's a way to get a good discussion, rather than a person giving a presentation15:35
edleafe2-way vs. 1-way15:35
edleafeIf it gets too heated, the session leader can call on people so that only one at a time is speaking15:36
edleafeBut generally that's not needed often15:36
edleafeThe name comes from the circular arrangement of chairs, resembling a goldfish bowl15:36
edleafeDoes that help?15:37
openstackgerritJean-Emile DARTOIS proposed openstack/watcher-specs: Update the continuously optimization specs.
edleafeFrom the OpenStack website: "Those are open sessions to discuss a specific feature or issue we have to solve. They happen in large rooms organized in fishbowl style (concentric rings of chairs). People wanting to participate to the discussion should move to the inner rings."15:38
jed56edleafe: thanks you very much !15:39
edleafejed56: you're welcome!15:40
jed56edleafe: if you have some time, I will love a review from you :)15:40
jed56this is not finish yet but, i would to have some feedback to continue15:41
*** Kevin_Zheng has quit IRC15:41
edleafejed56: already have that open in a tab in my reviwe queue :)15:41
openstackgerritJean-Emile DARTOIS proposed openstack/watcher-specs: Update the continuously optimization spec.
*** michaelgugino has quit IRC15:46
*** acabot has quit IRC15:57
*** hvprash_ has joined #openstack-watcher15:57
*** hvprash has quit IRC16:00
*** michaelgugino has joined #openstack-watcher16:00
*** jinquan1 has quit IRC16:08
*** jinquan has joined #openstack-watcher16:20
*** michaelgugino has quit IRC16:45
*** wootehfoot has joined #openstack-watcher17:32
*** jed56 has quit IRC18:45
openstackgerritMichael Gugino proposed openstack/watcher: WIP: Implement goal_id and watcher_id into Audit api
*** dtardivel has quit IRC19:17
*** openstackgerrit has quit IRC19:33
*** openstackgerrit has joined #openstack-watcher19:33
*** Zucan has quit IRC20:01
*** thorst has quit IRC20:02
openstackgerritMichael Gugino proposed openstack/watcher: WIP: Implement goal_id and watcher_id into Audit api
*** thorst has joined #openstack-watcher20:05
sballe_edleafe: Hi :) I am looking at the role of cgroups from an orchestrator/Watcher point of view and I am wondering if nova supports cgroups20:35
sballe_do you know? I cannot find any documentation on nova and cgroups20:36
*** hvprash has joined #openstack-watcher20:40
*** hvprash_ has quit IRC20:40
*** thorst has quit IRC21:08
*** thorst has joined #openstack-watcher21:20
*** thorst has quit IRC21:25
*** thorst has joined #openstack-watcher21:32
*** thorst has quit IRC21:36
*** thorst has joined #openstack-watcher21:44
*** hvprash has quit IRC21:44
*** thorst has quit IRC21:48
*** thorst has joined #openstack-watcher21:50
edleafesballe_: no, nova assumes it has full use of the host, and relies on the hypervisor to provide the necessary isolation for the VMs21:53
*** thorst_ has joined #openstack-watcher21:54
*** thorst has quit IRC21:54
*** thorst_ has quit IRC21:59
*** openstackgerrit has quit IRC22:02
*** openstackgerrit has joined #openstack-watcher22:05
*** thorst_ has joined #openstack-watcher22:21
*** thorst_ has quit IRC22:25
*** wootehfoot has quit IRC22:45

Generated by 2.14.0 by Marius Gedminas - find it at!