14:02:21 #startmeeting watcher 14:02:22 Meeting started Wed Dec 14 14:02:21 2016 UTC and is due to finish in 60 minutes. The chair is acabot. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:02:23 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:02:25 The meeting name has been set to 'watcher' 14:02:28 hi 14:02:29 hi everyone 14:02:32 hi 14:02:32 hi 14:02:33 sorry to be late 14:02:35 hi 14:02:43 o/ morning 14:02:49 Hi 14:02:58 our agenda for today #link https://wiki.openstack.org/wiki/Watcher_Meeting_Agenda#12.2F14.2F2016 14:03:07 I am on my phone so typing is slow 14:03:33 sballe_ : :-) ok I will let you more than to 2 sec to answer 14:03:54 thx 14:04:01 #topic Announcements 14:04:02 my internet is back now :) 14:04:16 #info we had a couple of gate issues last days due to upper constraints 14:04:27 thx licanwei for fixing it 14:04:33 +1 14:04:39 and dtardivel for reviews ;-) 14:04:42 +1 14:04:56 #info Watcher mid-cycle will be held in Atlanta in February Wednesday 22nd & Thursday 23rd 14:05:05 #link http://www.openstack.org/ptg 14:05:19 it will be at the PTG event with other openstack team 14:05:34 you need to register on eventbrite https://www.eventbrite.com/e/project-teams-gathering-tickets-27549298694 14:05:36 who has signed up ? 14:05:48 This time i am also attend this time if visa is issued ;-) 14:05:49 we will be 2 from b-com dtardivel & I 14:05:51 hello! 14:06:00 hi alexchadin 14:06:11 alexchadin : do you plan to be in Atlanta ? 14:06:14 ok we are 3 from intel. me, pshedimb and chrisspencer 14:06:33 tmetsch : do you want to join ? 14:06:36 atuly: do you need anythign for your visa 14:06:49 i already got IL 14:06:55 acabot: I'd be interested for sure. will check. 14:07:02 acabot: I think no:( 14:07:14 depends on budget also ;-) 14:07:26 any chance for sponsership . 14:07:28 alexchadin : ok we will try to have a hangout session (I hope timeshift will be ok for you) 14:07:41 acabot: It would be great 14:07:53 atuly: not sure. have you looked at #link http://www.openstack.org/ptg 14:08:02 yes 14:08:14 acabot: what's the deadline for registration ? 14:08:21 ok we should be at least 6 (3 Intel, 2 b-com, jwcroppe for IBM) 14:08:36 and Walmart will probably be there as well 14:08:44 around 10 people would be great 14:08:50 yes I talked to hvprash_ and he was going ot be there 14:09:20 Yumeng : February 20th 14:10:06 FYI, you have to buy a $100 ticket for the PTG but it includes a full pass for the Boston summit 14:10:37 ok 14:10:47 lets move to reviews as we have many things today... 14:10:54 #action Review Action Items 14:11:03 #undo 14:11:04 Removing item from minutes: 14:11:10 #topic Review Action Items 14:11:16 Specs 14:11:35 Define grammar for workload characterization needs a new PS an answering previous comments 14:11:46 #link https://review.openstack.org/#/c/377100/ 14:11:54 I am looking at it with hvprash 14:12:04 #action hvprash sballe_ add a new PS for https://review.openstack.org/#/c/377100/ 14:12:12 Support Description For Dynamic Action needs completion 14:12:23 #link https://review.openstack.org/#/c/401111/ 14:12:30 yes, I will 14:12:33 thx 14:12:45 and hello hanrong1 ;-) 14:12:56 :) 14:12:58 #action hanrong1 complete spec https://review.openstack.org/#/c/401111/ 14:13:05 Add specs for cache occupancy strategy needs a new PS 14:13:16 Add specs for cache occupancy strategy needs a new PS 14:13:27 #link https://review.openstack.org/#/c/398162/ 14:14:22 yeah pshedimb and I working on it. 14:14:27 sballe_ : I'm concerned about this spec because the associated BP has 2 BPs dependancies that have not been implemented yet 14:14:48 I'm really not sure we can have it in Ocata 14:15:13 acabot: then let's undo the dependencies. Because right now we are looking at the algorithms and pshedimb will start implementing so we can iterate over the algorithm 14:15:27 I hoped that characterization grammar would have been done earlier 14:15:45 the issue here is thta when we put out the spec we called it cache occupancy and it should really be noisy neighbor strategy 14:16:10 sballe_ : why dont you rename it ? 14:16:21 sballe_ : it would be more readable 14:16:31 ok will do. 14:16:40 lets discuss dependencies in open discussions 14:16:48 Audit tag in VM Metadata needs final reviews from core 14:16:59 #link https://review.openstack.org/#/c/391374/ 14:17:08 and it also ties much closer to the QoS BP so we might just replace the QOS BP with the noisy neightbor BP 14:17:31 #action sballe_ jwcroppe hvprash review https://review.openstack.org/#/c/391374/ 14:18:02 Add improvements to the planner and workflow mechanisms needs a jenkins fix & reviews 14:18:08 #link https://review.openstack.org/#/c/385871/ 14:18:23 alexchadin : do you need help with jenkins fix ? 14:18:40 alexchadin: when you do you will be my hero :) 14:18:54 jenkins has been a pain 14:19:07 acabot: I will make it myself:) 14:19:23 acabot: Just have a little time because of end of year 14:19:32 thx 14:19:46 alexchadin: if you need help I believe pshedimb can help 14:19:47 #action alexchadin fix jenkins for https://review.openstack.org/#/c/385871/ 14:19:58 sballe_: ok, thanks! 14:19:58 who wants to review this important spec ? 14:20:10 the jenkins fix? 14:20:19 after jenkins fix :-) 14:20:32 #link https://review.openstack.org/#/c/385871/ 14:20:33 acabot: I've started to write Applier spec, hope it will be available to review by EOW 14:20:59 alexchadin : great, I moved it to ocata-3 scope 14:21:08 acabot: ok, cool 14:21:18 moving to Watcher code 14:21:28 ocata-2 release is due tomorrow 14:21:35 so we have a lot of code to merge 14:21:43 centralise-config-opts must be merged 14:21:54 #link https://review.openstack.org/#/q/status:open+project:openstack/watcher+branch:master+topic:bp/centralise-config-opts 14:21:56 on going got this one 14:21:59 acabot: I will review it ASAP 14:22:01 for this one 14:22:28 thx alexchadin, dtardivel is already merging them one by one 14:22:35 I need to rebase locally a lot of changesets. I will take time because gate is slow ... 14:22:51 update strategy table when parameters_spec changes must be merged 14:22:57 #link https://review.openstack.org/#/c/408411/ 14:23:13 alexchadin can you W+1 ? 14:23:16 alexchadin: can you merge this one ? 14:23:33 dtardivel: acabot: ok 14:23:41 #action alexchadin review & W+1 https://review.openstack.org/#/c/408411/ 14:23:51 Add auto_trigger support to watcher needs final reviews & merge 14:24:01 #link https://review.openstack.org/#/c/403718/ 14:24:04 but after I finish with config-opt please .... :) 14:24:31 #action dtardivel review and merge https://review.openstack.org/#/c/403718/ 14:25:06 New default planner needs final reviews & merge 14:25:15 #link https://review.openstack.org/#/c/406991/ 14:25:33 +1 14:25:51 alexchadin : can you fix licanwei comment ? 14:26:10 before dtardivel starts reviewing 14:26:37 acabot: I will do till end of this day 14:26:42 thx 14:26:51 acabot: a lot of work, sorry 14:27:03 #action alexchadin fix comment for https://review.openstack.org/#/c/406991/ 14:27:16 check the state of action plan needs final reviews & merge 14:27:22 #link https://review.openstack.org/#/c/391383/ 14:27:46 who wants to review it ? 14:27:51 +1 14:27:53 * exploreshaifali wants to 14:28:03 +1 14:28:10 #action Yumeng exploreshaifali review https://review.openstack.org/#/c/391383/ 14:28:22 +1 14:28:30 #action dtardivel review & merge https://review.openstack.org/#/c/391383/ 14:28:38 #action atuly review https://review.openstack.org/#/c/391383/ 14:28:57 please keep in mind that we need your reviews before tomorrow EOD 14:29:09 okay 14:29:11 ok 14:29:15 Repair log parameter error needs a bug reference & reviews 14:29:24 #link https://review.openstack.org/#/c/407901/ 14:29:59 #action licanwei submit a new bug and add bug ID to patch https://review.openstack.org/#/c/407901/ 14:30:05 who wants to review it ? 14:30:15 I think bug is already there 14:30:22 * exploreshaifali wants to 14:30:35 atuly : bug # ? 14:30:50 sorry got confused with other review.. 14:30:51 #action exploreshaifali review https://review.openstack.org/#/c/407901/ 14:30:58 atuly : ok ;-) 14:31:13 #action dtardivel review and merge https://review.openstack.org/#/c/407901/ 14:31:24 Watcher CLI 14:31:31 Add auto_trigger support needs reviews & merge 14:31:37 #link https://review.openstack.org/#/c/403716/ 14:32:08 +1 14:32:20 #action dtardivel review & merge https://review.openstack.org/#/c/403716/ 14:32:29 Support parents field along with planner changes needs reviews & merge 14:32:36 #link https://review.openstack.org/#/c/409632/ 14:32:46 opps. sorry.. I may not be able to review https://review.openstack.org/#/c/391383/ before tomorrow EOD.. already late night in my time.. 14:32:50 already merged :-) 14:32:57 ll 14:32:59 lol 14:33:12 Yumeng : ok no pb 14:33:51 wrong link...let me check 14:33:54 #undo 14:33:55 Removing item from minutes: #link https://review.openstack.org/#/c/409632/ 14:34:22 #link https://review.openstack.org/#/c/409126/ 14:34:25 sorry 14:34:29 who wants to review ? 14:34:48 this patch needs reviews before being merged tomorrow 14:34:54 +1 14:34:59 +1 14:35:03 14 lines ;-) 14:35:07 thx 14:35:12 haha 14:35:22 #action exploreshaifali atuly review https://review.openstack.org/#/c/409126/ 14:35:31 #action dtardivel review and merge https://review.openstack.org/#/c/409126/ 14:36:06 dtardivel has a lot of merge to do so please try to help him with +1 reviews 14:36:16 yep, thx 14:36:29 #topic Blueprint/Bug Review and Discussion 14:36:32 i may be able to review only be tomrw morning.. 14:36:54 atuly : tomorrow morning your time is fine 14:37:05 okay 14:37:18 atuly : we will release tomorrow around 2pm GMT 14:37:30 same time as the current meeting 14:37:40 ok :) 14:37:48 #info Watcher ocata-2 must be released tomorrow December 15th 14:37:54 #link https://launchpad.net/watcher/+milestone/ocata-2 14:38:18 I have moved all unimplemented BPs to ocata-3 14:38:53 Can I ask a question? I have two doc spec and people keep on adding stuff to the comments. I am happy to oblige but will need a +2 after people have given it +1s 14:39:02 We are targeting 4 BPs & 7 bugs fixed 14:39:49 sballe_: links ? 14:40:10 #link https://review.openstack.org/404960 & https://review.openstack.org/404972 14:40:24 https://review.openstack.org/#/c/404972/6 and https://review.openstack.org/#/c/404960/3 14:40:27 sballe_ : thats right, we need to merge them 14:41:01 #action acabot dtardivel review and merge https://review.openstack.org/404972 & https://review.openstack.org/404960 14:41:02 I'fix the latest comments and then let's merge them. for example some of the comments are true for past doc as well 14:41:30 sballe_ : please add a new PS asap, I will review it by EOD 14:41:44 yes but after you take into account last -1 comment :) please 14:41:44 I will do it withing ht enext hour 14:41:52 thx 14:41:55 #info Improve Applier to execute Actions in parallel has been added 14:42:00 dtardivel: of course. 14:42:06 #link https://blueprints.launchpad.net/watcher/+spec/parallel-applier 14:42:12 sballe_: please ping me if you need help 14:42:18 as alexchadin has been splitted in 2 14:42:27 :-) 14:42:36 alexchadin BP has been splitted 14:42:40 the issue is that if you go back into the older doc the same comment-1 is tru for some of those. how do we want to deal with that 14:42:42 :D 14:42:43 dtardivel: ^ 14:43:40 sballe_ : please move this discussion to #openstack-watcher 14:43:46 ok 14:43:54 is vm-migration-ordering dependent on Improve Applier to execute Actions in parallel ? 14:44:05 #link https://blueprints.launchpad.net/watcher/+spec/vm-migration-ordering 14:44:11 #link https://blueprints.launchpad.net/watcher/+spec/parallel-applier 14:44:41 alexchadin : I think we have a dependency here with the new BP you submitted, what do you think ? 14:45:55 acabot: seems that this BP propose to set actions order by strategies 14:46:18 action's* 14:47:06 acabot: looks like dependency, yes 14:47:06 alexchadin : so does it mean we should keep them separated ? 14:47:18 acabot: yes, absolutely 14:47:31 I think VM migration ordering would be dependant from your BP 14:47:49 acabot: I need more info about these proposing things 14:47:58 acabot: yes 14:48:21 ok dependency added 14:48:38 Strategy to detect and migrate the the cache greedy VM affecting other's performance is dependent on Define grammar for workload characterization & Workload Characterization and QoS which have not been started yet, I would suggest to postpone it to Pike cycle 14:49:04 sballe_ : should we get rid of dependencies in https://blueprints.launchpad.net/watcher/+spec/cache-occupancy-strategy ? 14:49:23 yes 14:49:43 sballe_ : if yes I would suggest to rename it "noisy neighbour strategy" 14:49:50 to avoid confusion 14:49:57 before removing dependancies 14:50:03 dependencies 14:50:29 I will do that and the noisy neighbor strategy is very much the Workload characteristic and QoS BP 14:51:08 We have decided to craete one useful strategy which is why we are doign it 14:51:11 this way 14:51:30 but for now let's keep noisy neightbor and the otehr one 14:51:33 sballe_ : ok so what would be the diff between https://blueprints.launchpad.net/watcher/+spec/cache-occupancy-strategy and https://blueprints.launchpad.net/watcher/+spec/workload-characterization-qos then ? 14:52:12 for now let's keep both but as I work throught hw details they migth end up being the same 14:52:59 sballe_ : ok :-D very easy for me to plan it for ocata in these conditions 14:53:38 sorry ;-( 14:53:42 sballe_ : I suppose you can at least remove the dependency with characterization grammar bp 14:53:56 yes 14:54:14 I would like to work on "characterization grammar" BP, if it is needed 14:54:20 and I will do that as soon as I rename the https://blueprints.launchpad.net/watcher/+spec/cache-occupancy-strategy BP 14:54:23 #action sballe_ review & rename BPs https://blueprints.launchpad.net/watcher/+spec/cache-occupancy-strategy & https://blueprints.launchpad.net/watcher/+spec/workload-characterization-qos and remove dependency with https://blueprints.launchpad.net/watcher/+spec/workload-characterization-grammar 14:54:45 exploreshaifali: you and hvprash_ are already workingon it :-) 14:55:06 yes, but not started writing code 14:55:11 exploreshaifali : you are more than welcome to add a PS or reviews to https://review.openstack.org/#/c/377100/ 14:55:33 okay. Thanks :) 14:55:53 #topic Open discussions 14:56:04 I don't have any topic today 14:56:13 4 minutes left 14:56:44 again we have to release tomorrow so please keep the focus on "in review BPs" 14:56:57 neither do I. Do we have a IRC meeting on 12/28? 14:57:14 acabot: will do 14:57:30 good point, I will be off for the next 2 weeks but vincentfrancoise will be there 14:57:54 do you want to keep weekly meetings during xmas holidays ? 14:58:08 ok. I will be off the week between x-mas and new year 14:58:10 I will ask vincent to handle it if needed 14:58:38 acabot: do you want to discuss https://blueprints.launchpad.net/watcher/+spec/cluster-energy-saving next week? or we can discuss on openstack-watcher ? 14:58:56 ok, vincent will open the meeting as usual and if there are not enough people it will be canceled 14:58:59 before we start spec 14:59:12 #action vincentfrancoise open meetings during xmas holidays 14:59:16 Yumeng: I am interseted int his topic as well. please ping me if I am around 14:59:33 sballe_: no pb 14:59:42 Yumeng : sorry I missed it 14:59:54 I will add it to the agenda next week 15:00:05 acabot: no worries. thx :) 15:00:15 have a good day/night 15:00:19 bye 15:00:23 bye 15:00:25 bye 15:00:26 bye! 15:00:33 #endmeeting