Wednesday, 2016-05-18

*** hvprash has joined #openstack-watcher00:30
*** hvprash has quit IRC00:34
*** jwcroppe has joined #openstack-watcher01:08
*** jwcropp__ has joined #openstack-watcher01:09
*** jwcroppe_ has quit IRC01:10
*** jwcroppe has quit IRC01:12
*** hvprash has joined #openstack-watcher01:55
*** hvprash has quit IRC02:00
*** thorst_ has joined #openstack-watcher02:13
*** thorst_ has quit IRC02:36
*** thorst_ has joined #openstack-watcher02:36
*** thorst_ has quit IRC02:41
*** xenogear has quit IRC04:30
*** vmahe1 has joined #openstack-watcher05:04
*** jwcroppe has joined #openstack-watcher05:09
*** jwcropp__ has quit IRC05:12
*** ecelik has quit IRC05:12
*** vmahe has quit IRC05:12
*** ecelik has joined #openstack-watcher05:13
*** apoorv has joined #openstack-watcher06:08
apoorvgetting this error on running audit-template-list : Unknown attribute for argument audit_template_postdata: goal (HTTP 400)06:21
jed56sballe_: we have on video but she is old06:59
jed56apoorv: IMHO, you have to update the watcher client07:16
*** vincentfrancoise has joined #openstack-watcher07:19
*** apoorv has quit IRC07:45
*** Kevin_Zheng_ has joined #openstack-watcher08:42
*** akijak_ has joined #openstack-watcher08:45
*** tpeoples has quit IRC08:45
*** Kevin_Zheng has quit IRC08:45
*** ecelik has quit IRC08:45
*** akijak has quit IRC08:45
*** Kevin_Zheng_ is now known as Kevin_Zheng08:46
*** ecelik has joined #openstack-watcher08:46
*** tkaczynski has quit IRC08:48
*** tpeoples has joined #openstack-watcher08:52
*** dtardivel has joined #openstack-watcher08:55
openstackgerritAlexander Chadin proposed openstack/watcher-specs: Add Watcher Continuosly Optimization spec
*** tkaczynski has joined #openstack-watcher09:02
*** Kevin_Zheng has quit IRC09:16
*** apoorv has joined #openstack-watcher09:19
*** Kevin_Zheng has joined #openstack-watcher09:21
*** alexchadin has joined #openstack-watcher09:33
openstackgerritDavid TARDIVEL proposed openstack/python-watcherclient: OpenStackClient plugin for audit template
openstackgerritDavid TARDIVEL proposed openstack/python-watcherclient: OpenStackClient plugin for action plan
*** zarkone has joined #openstack-watcher10:08
zarkoneHello all! Have another problem: there is no metric "compute.node.cpu.percent" in ceilometer... tried this solution, but seems like it's not my case:
*** hvprash has joined #openstack-watcher10:30
*** hvprash has quit IRC10:35
*** akijak_ is now known as akijak10:36
*** akijak has left #openstack-watcher10:37
*** akijak has joined #openstack-watcher10:37
alexchadinhi zarkone!10:43
alexchadinare you still here?10:44
zarkonealexchadin: yep ;) Hi!10:44
alexchadinzarkone: you need to set compute_monitors to cpu.virt_driver on your compute nodes in nova configuration file10:45
alexchadinafter you restart nova-compute service this metric should be added to ceilometer samples10:45
zarkonealexchadin: ok, give me a few minutes, i'll check it again10:46
openstackgerritDougal Matthews proposed openstack/python-watcherclient: Use the correct capitalization of OpenStack
openstackgerritDougal Matthews proposed openstack/python-watcherclient: Use the correct capitalization of OpenStack
*** AndChat|176409 has joined #openstack-watcher10:48
*** alexchadin has quit IRC10:48
*** AndChat|176409 has quit IRC10:49
*** alexchadin has joined #openstack-watcher10:49
zarkonealexchadin: so, i went to one of the compute nodes, set compute_monitors and restarted nova-compute10:50
zarkone ceilometer sample-list | awk '{print $6}' | grep cpu10:50
zarkonecpu_util, cpu, cpu.delta10:51
zarkonethat's all happen on Liberty10:53
alexchadinThat's strange10:53
alexchadinDid you check ceilometer logs?10:53
alexchadinCollector logs10:55
zarkone cat /var/log/ceilometer/agent-notification.log | grep cpu10:56
zarkone2016-05-17 23:52:18.122 25441 WARNING ceilometer.transformer.conversions [req-e14cc7b6-2b4f-4790-96be-ad5efaf815b5 - - - - -] Dropping sample with no predecessor: (<name: cpu, volume: 6870000000, resource_id: 0bbb9497-2c27-4d7c-bf01-df5f4b115011, timestamp: 2016-05-18T04:52:18.109090>,)10:56
alexchadinMaybe you have to restart ceilometer service on compute node10:56
zarkoneI tried to restart ceilometer-polling10:58
zarkoneand then looked at logs10:59
zarkonea lot of errors10:59
zarkonethanks, alexchadin, your advice moved my situation10:59
*** AndChat|176409 has joined #openstack-watcher10:59
zarkonewill debug this errors now10:59
*** AndChat|176409 has quit IRC11:00
*** AndChat|176409 has joined #openstack-watcher11:00
*** alexchadin has quit IRC11:01
*** AndChat|176409 has quit IRC11:01
*** alexchadin has joined #openstack-watcher11:01
alexchadinSorry for disconnecting11:02
alexchadinWhat did you say?11:02
zarkone17:59 zarkone: I tried to restart ceilometer-polling11:03
zarkone17:59 zarkone: and then looked at logs11:03
zarkone17:59 zarkone: a lot of errors11:03
zarkone17:59 zarkone: thanks, alexchadin, your advice moved my situation11:03
zarkone17:59 zarkone: will debug this errors now11:03
alexchadinTry to monitor ceilometer collector service, there should be logs of all gathered metrics11:05
*** Kevin_Zheng has quit IRC11:11
*** alexchadin has quit IRC11:17
*** thorst_ has joined #openstack-watcher11:35
*** thorst__ has joined #openstack-watcher11:35
*** thorst_ has quit IRC11:39
*** alexchadin has joined #openstack-watcher12:08
apoorvjed56: not able to resolve that error12:08
vincentfrancoisehi apoorv12:09
vincentfrancoisesince the get-goal-from-strategy blueprint12:09
vincentfrancoisethe --goal parameter of the CLI and goal attribute of the audit template has been renamed respectively as --goal-uuid and goal_uuid12:10
vincentfrancoisethere is an gerrit changeset to change back the CLI to now accept --goal as the goal UUID or the goal name BTW12:11
*** apoorv has quit IRC12:20
openstackgerritMerged openstack/watcher: Watcher DB class diagram
alexchadinvincentfrancoise: hi:)12:37
vincentfrancoisealexchadin: hi12:37
alexchadinI'm testing constinuously-optimization right now12:37
alexchadinIt works well12:37
vincentfrancoiseah good :)12:38
alexchadinWe also already made changes for CLI12:38
vincentfrancoisedid you do it the magnum way?12:38
alexchadinadapted it, yeah12:38
vincentfrancoisehow did you do it for the CLI12:39
vincentfrancoiseon top of the OSC refactoring12:39
vincentfrancoiseor the old way?12:39
alexchadinjust added --period support for audit-create command and next_launch, period fields to audit-show12:40
alexchadinWe've made it for watcherclient12:40
alexchadinnot for OSC12:40
vincentfrancoiseyeah but in the blueprint I just linked, I actually refactored the watcherlient CLI to be used both as an OSC client and as a watcher command12:42
vincentfrancoisebut if it's not too much changes, it shouldn't be a problem to upgrade anyway ;)12:42
alexchadinVladimir Ostroverkhov will looking through12:43
vincentfrancoiseWho is he? one of you colleague?12:43
alexchadinYes, he is junior engineer and works with me12:44
alexchadinI'm his mentor for now12:44
vincentfrancoiseoh ok I see12:44
alexchadintrying to involve him to contributing12:44
vincentfrancoisehe can come and ask me questions if he needs some help ;)12:44
alexchadinI'll tell him:)12:45
vincentfrancoisewhat's his irc nickname?12:45
alexchadinAnd sorry, I won't be on meeting today, but he will12:45
alexchadinProbably xrolla12:45
alexchadinHe didn't logged in12:45
vincentfrancoiseok will tell acabot ;)12:45
vincentfrancoiseare you going to push your changes before the meeting?12:46
*** Kevin_Zheng has joined #openstack-watcher12:46
alexchadinI actually didn't run any tests...12:49
alexchadinSo I have to adapt them and make new ones12:49
vincentfrancoiseoh ok fair enough, good luck with that :p12:50
alexchadinI'll need it, thank you :)12:51
alexchadinBut spec is already pushed and waiting for reviews12:52
vincentfrancoisejed56 is reviewing your strategy at the moment12:52
vincentfrancoiseso maybe he'll carry on making reviews on this one12:53
vincentfrancoiseI'll try to read it before EOW12:54
alexchadinIt would be great12:54
vincentfrancoiseI'm currently focusing on finishing my blueprint on efficacy-indicator12:55
alexchadinYou want me to review it?12:55
vincentfrancoisedidn't push the code yet12:55
vincentfrancoisebut probably tomorrow12:55
vincentfrancoisethanks anyway12:56
alexchadinAdd me to reviewers to get me notificated12:56
vincentfrancoiseok I will add you then ;)12:56
alexchadinonce you pushed it12:56
*** thorst__ is now known as thorst_13:17
*** hvprash has joined #openstack-watcher13:18
*** hvprash_ has joined #openstack-watcher13:21
*** hvprash has quit IRC13:24
alexchadinjed56: i have a question about your comments13:27
jed56alexchadin: I'm currently in a meeting13:27
jed56but feel free to ask them :)13:27
alexchadinjed56: about computationally expensive - for this reason i use memcached to store all received data from ceilometer till the end of strategy work13:29
alexchadinis it bad pattern?13:32
*** lrensing has joined #openstack-watcher13:37
*** alexchadin has quit IRC13:38
*** hvprash_ has quit IRC13:45
*** hvprash has joined #openstack-watcher13:45
*** ostroverkhov has joined #openstack-watcher13:57
jed56alexchadin : the problem is that you are exploring every possible solution13:58
*** jwcroppe has quit IRC13:59
acabotweekly meeting starts on #openstack-meeting-413:59
*** jwcroppe has joined #openstack-watcher13:59
*** ostroverkhov is now known as xrollla14:00
*** brunograz has joined #openstack-watcher14:00
*** vmahe1 has quit IRC14:02
*** ecelik has quit IRC14:28
*** jwcroppe_ has joined #openstack-watcher14:51
*** jwcroppe has quit IRC14:53
*** jwcroppe has joined #openstack-watcher14:57
*** jwcroppe_ has quit IRC15:00
*** jwcroppe_ has joined #openstack-watcher15:00
sballe_gzhai2: acabot let's continue the cluster scale discussion15:00
gzhai2What I need is the requirements and timeline.15:00
jed56 what are the characteristics of the nodes ?15:01
sballe_gzhai2: can get access tto 30+ nodes15:01
*** Kevin_Zheng has quit IRC15:01
acabotgzhai2 : do you have a template that we can fill ?15:01
acabotgzhai2 : what level of details do you need on requirements ?15:01
gzhai2like from when to when?15:02
sballe_acabot: one of the ask is for how long do we need it. gzhai2 will need some guidance around how to run or debug any issues15:02
sballe_so let's say for example we say we need it all of June15:03
acabotgzhai2 : do you have a tool to "stress" the environment ?15:03
*** jwcroppe has quit IRC15:03
acabotI mean something to simulate workloads in VMs15:03
gzhai2acabot: no.15:03
sballe_we will need to come up with a test plan15:03
gzhai2we can find it.15:03
gzhai2yes. We need test plan.15:03
gzhai2most important is time line.15:04
openstackgerritMerged openstack/python-watcherclient: Use the correct capitalization of OpenStack
gzhai2how long would we use it.15:04
sballe_lets make sure we have everyhting lined up so we do not waste time looking for tools after we have the cluster. could we setup an etherpad to describe steps and stuff we want to do15:04
acabotI think 1 month would be great15:04
acabotsballe_ : sure15:05
gzhai2acabot: from when?15:05
gzhai2I need submit request before May20 for internal cluster.15:06
sballe_acabot: When will your team have the time to help gzhai2 ? shoudl we ask for June 15-July 15 that way we have a little time to prep15:07
acabotgzhai2 : I think we should run scalability tests between Newton-3 & RC-final (August 29th - September 26th)15:07
sballe_acabot: +1 make a lot of sense to me15:07
sballe_acabot: Can we narow this down to one month?15:08
acabotI think september would be a good fit (my team will be available)15:08
gzhai2I'll submit request for September15:08
sballe_gzhai2: thx15:09
gzhai2And we can make test plan before it.15:09
acabotbut we need to have a 30 nodes cluster, running OpenStack with Telemetry/monasca deployed to focus on Watcher deployment (not on OpenStack issues)15:09
sballe_gzhai2: can you also get the specs for the servers?15:09
*** dronshaw has joined #openstack-watcher15:09
sballe_acabot: let's put that down on the etherpad and gzhai2 and I will make sure we have the dependencies etup before hand15:10
sballe_it is good we have 2 months to get this organized15:10
gzhai2acabot: will ask spec.15:10
sballe_cool! thx15:12
*** lrensing has quit IRC15:12
acabotthx to both of you15:12
gzhai2thank you15:13
*** xrollla has quit IRC15:23
openstackgerritVincent Françoise proposed openstack/watcher: Decoupled Goal from Strategy
openstackgerritVincent Françoise proposed openstack/watcher: Added EfficacyIndicator model in DB
openstackgerritVincent Françoise proposed openstack/watcher: Added efficacy specification to /goals
openstackgerritVincent Françoise proposed openstack/watcher: Added EfficacyIndicator object
openstackgerritVincent Françoise proposed openstack/watcher: Added efficacy indicators to /action_plans
edleafeacabot: Updated the commit message etherpad15:48
*** brunograz has quit IRC16:02
*** zarkone has quit IRC16:03
*** hvprash_ has joined #openstack-watcher16:05
*** hvprash has quit IRC16:08
*** vincentfrancoise has quit IRC16:15
*** alexchadin has joined #openstack-watcher16:57
alexchadinedleafe: hi16:57
alexchadinedleafe: thank you for your review, i have some questions16:58
edleafealexchadin: sure16:58
alexchadinedleafe: regarding 'continuous' name - when i started to work on this project there already was oneshot and continuous types of audit16:59
alexchadinedleafe: i think we have to ask other members of changing continuous to periodic17:01
alexchadinacabot: edleafe proposed to change continuous type to periodic17:02
edleafealexchadin: I don't have the same history, so when I read your spec, 'continuous' didn't sound right17:02
alexchadinedleafe: i got you17:02
edleafeTo me, that just means 'constantly running'17:02
*** hvprash has joined #openstack-watcher17:03
alexchadinok thank you17:04
alexchadinWe can specify frequency of periodic_time's calling by Watcher configuration file.17:04
alexchadinI forgot to remove it17:04
alexchadinWe have 'period' field in audit table17:04
*** dronshaw has quit IRC17:04
alexchadinWe can specify period for every audit we are creating17:05
*** hvprash_ has quit IRC17:06
*** hvprash has quit IRC17:18
*** hvprash has joined #openstack-watcher17:19
*** alexchadin has quit IRC17:26
*** hvprash_ has joined #openstack-watcher17:29
*** hvprash has quit IRC17:32
*** dtardivel has quit IRC18:00
sballe_hi does anybody remember the global scheduler BP that I believe Jerry Harrow wrote up? I cannot find it18:16
*** hvprash_ has quit IRC18:43
*** hvprash has joined #openstack-watcher18:43
*** hvprash_ has joined #openstack-watcher18:53
*** hvprash has quit IRC18:55
*** jwcroppe has joined #openstack-watcher20:20
*** jwcroppe_ has quit IRC20:22
openstackgerritOpenStack Proposal Bot proposed openstack/python-watcherclient: Updated from global requirements
*** thorst_ has quit IRC21:12
*** thorst_ has joined #openstack-watcher21:16
*** thorst_ has quit IRC21:20
*** thorst_ has joined #openstack-watcher21:34
*** thorst_ has quit IRC21:38
*** hvprash_ has quit IRC22:01
*** hvprash has joined #openstack-watcher22:02
*** jwcroppe_ has joined #openstack-watcher22:17
*** jwcroppe has quit IRC22:20
*** harlowja has quit IRC22:22
*** hvprash_ has joined #openstack-watcher22:36
*** hvprash has quit IRC22:39
*** harlowja has joined #openstack-watcher23:00
*** hvprash_ has quit IRC23:22

Generated by 2.14.0 by Marius Gedminas - find it at!