Wednesday, 2016-05-04

*** thorst has joined #openstack-watcher01:09
*** thorst has quit IRC01:57
*** thorst has joined #openstack-watcher01:58
*** thorst has quit IRC02:06
*** thorst has joined #openstack-watcher03:04
*** thorst has quit IRC03:11
*** Kevin_Zheng has joined #openstack-watcher03:17
*** thorst has joined #openstack-watcher04:09
*** thorst has quit IRC04:16
*** vtech has quit IRC04:35
*** vtech has joined #openstack-watcher04:36
*** apoorv has joined #openstack-watcher05:01
*** apoorv has quit IRC05:03
*** vtech has quit IRC05:03
*** vtech has joined #openstack-watcher05:05
*** apoorv has joined #openstack-watcher05:10
*** thorst has joined #openstack-watcher05:14
*** thorst has quit IRC05:21
*** openstackgerrit has quit IRC06:17
*** openstackgerrit has joined #openstack-watcher06:17
*** thorst has joined #openstack-watcher06:18
*** thorst has quit IRC06:26
*** vtech has quit IRC06:35
*** esracelik has joined #openstack-watcher06:58
*** vtech has joined #openstack-watcher06:59
*** vtech_ has joined #openstack-watcher06:59
*** vtech has quit IRC06:59
*** thorst has joined #openstack-watcher07:24
apoorvjed56, thanks that helps. However, i was just wondering if we have to pack VMs on one single node, why is CPU utilization used as a metric to rank the individual hosts?07:24
apoorvjed56, is this a way of telling, that a given host is the most loaded (even if it has fewer VMs) - and we should migrate to it?07:27
jed56hello apoorv07:30
apoorvI am talking about a scenario where let's say a given host has 1 VM - but that has hogged the resources so the CPU utilization is high on the node but then there is another node that has 5 smaller idle VMs - hence the node utilization is low. But if we sort with the utilization - then the first node will get all the other VMs on instead of migrating just oneVM - we will migrate 4.07:30
*** thorst has quit IRC07:31
jed56apoorv: it is a limitation of basic consolidation07:32
jed56we only use CPU because by default ceilometer have these metrics07:32
jed56we should not only CPU for rank an host07:33
jed56moreover, we need to use predicted usage of the VMS when is possible to avoid oscillation07:34
*** alexchadin has joined #openstack-watcher07:43
*** apoorv has quit IRC07:51
alexchadinHello everyone!08:20
alexchadinDid you have a discussion of this on summit?
*** akijak has joined #openstack-watcher08:22
jed56alexchadin: you can ask acabot08:24
jed56during the meeting08:24
alexchadinjed56: yeap:)08:24
jed56alexchadin: the blueprint is unassigned08:24
alexchadinI see08:25
jed56so, you can probably take it08:25
vincentfrancoisethe essential ones are the BPs we want to implement before joining the big tent08:25
jed56I suggest you to add a point08:26
jed56in the agenda08:26
vincentfrancoiseso I believe most the talks were on these08:26
vincentfrancoiseso I don't think you and I missed anything important regarding this BP08:26
alexchadinI want to discuss some drafts regarding this BP on the meeting08:27
vincentfrancoisethe only mention to this BP is about tpeoples being interested to take it08:27
alexchadinOkay, let's take a look on this on meeting08:28
*** thorst has joined #openstack-watcher08:29
alexchadin"IOPS seems to an interesting metric to work on after CPU & RAM"08:29
acabothi alexchadin08:29
*** akijak has quit IRC08:29
acabotwe decided to set 1 essential BP per contributor to Watcher for N1 release08:29
alexchadinI'll test sd-strategy with IOPS08:29
acabotso I can move continuous optimization to essential if you plan to work on it in May08:30
acabotbut I need to check that sd-strategy can now be merged08:30
alexchadinacabot: I do08:30
acabotcan you add something in the blueprints section of the agenda ?08:31
acabotto tell that you would like to have this blueprint for Servionica08:31
alexchadinIf everything seems to be good I will be writing spec tomorrow.08:32
jed56alexchadin: great :)08:35
*** thorst has quit IRC08:36
*** akijak has joined #openstack-watcher08:38
*** sudipto has joined #openstack-watcher08:39
alexchadinacabot: Done08:42
*** apoorv has joined #openstack-watcher08:46
*** thorst has joined #openstack-watcher09:33
*** thorst has quit IRC09:41
*** pav0 has joined #openstack-watcher10:02
*** pav0 has left #openstack-watcher10:03
*** pav0 has joined #openstack-watcher10:03
pav0Hello, I want to create puppet module for watcher. Could someone tell me, which license file should I include:   from watcher module (there is Apache 2.0 license  ) or from another puppet module like puppet-nova ( there is Copyright 2012 OpenStack Foundation) ?10:14
*** vincentfrancoise has quit IRC10:15
*** thorst has joined #openstack-watcher10:39
*** thorst has quit IRC10:46
sudiptojed56, Hi, w.r.t apoorv 's question - shouldn't scoring the node also take into account the number of VMs running on it?10:49
*** alexchadin has quit IRC11:15
*** alexchadin has joined #openstack-watcher11:15
*** vtech_ has quit IRC11:27
*** vtech has joined #openstack-watcher11:28
*** alexchadin has quit IRC11:29
*** sudipto has quit IRC11:41
jed56sudipto hello11:47
jed56yes this is a good idea11:47
jed56pav0: hello11:50
jed56pav0: I believe that the OpenStack project are under the Apache 2.0 license11:51
*** thorst has joined #openstack-watcher11:51
*** apoorv has quit IRC12:07
pav0jed56: hello, thanks. Both files are based on Apache 2.0, but in one file ( Copyright are for OpenStack Foundation and in watcher LICENSE file (  there is no mentioning about copyright for OpenStack foundation12:10
jed56good to know12:10
pav0jed56: so I guess I should take LICENSE file from other puppet-<OPENSTACK PROJECT> project :P12:19
jed56I think so, what do you think acabot ?12:20
*** vincentfrancoise has joined #openstack-watcher12:22
acabotpav0 : yes I think its the best way to solve this, I look take a look at our licences consistency next week12:22
acabotI will take a look12:22
pav0acabot: thanks12:22
*** pav0 has left #openstack-watcher12:24
*** pav0 has joined #openstack-watcher12:24
pav0one more question: should I create new project on Launchpad like puppet-nova, puppet-cinder etc. or someone of "core" watcher team can create it?12:26
vincentfrancoisepav0: I don't know, maybe we can put this to today's meeting's agenda12:30
pav0vincentfrancoise: it will be great :) thanks12:31
vincentfrancoisepav0: done12:33
*** hvprash has joined #openstack-watcher13:24
*** sudipto has joined #openstack-watcher13:24
*** hvprash_ has joined #openstack-watcher13:27
*** hvprash has quit IRC13:28
*** hvprash has joined #openstack-watcher13:29
openstackgerritVincent Mahe proposed openstack/watcher-specs: Provide efficacy indicators
*** hvprash_ has quit IRC13:32
*** sudipto has quit IRC13:36
*** dronshaw has joined #openstack-watcher13:56
*** alexchadin has joined #openstack-watcher13:57
acabothi guys, our weekly meeting will start in 2 min on #openstack-meeting-413:58
dronshawthanks for reminding :)13:59
*** jwcroppe has joined #openstack-watcher14:01
*** hvprash has quit IRC14:11
*** hvprash has joined #openstack-watcher14:12
*** esracelik has quit IRC14:14
*** vincentfrancoise has quit IRC14:15
*** vincentfrancoise has joined #openstack-watcher14:16
*** jwcroppe has quit IRC14:23
*** vtech has quit IRC14:30
*** jwcroppe has joined #openstack-watcher14:31
*** jwcroppe_ has joined #openstack-watcher14:32
*** jwcroppe has quit IRC14:36
*** jwcroppe_ has quit IRC14:44
*** jwcroppe has joined #openstack-watcher14:45
*** vtech has joined #openstack-watcher14:56
jed56pav0: we can start by the main variables first14:59
sballe_jwcroppe can you attend the congress meeting tomorrow?15:00
jed56in order to have  the "minimal"15:00
sballe_jwcroppe: acabot has the details15:00
pav0jed56: I will try to do as much as possible15:00
sballe_jwcroppe: acabot we just need to make sure we know what is going on and in case we get some follow-ups we should be thre15:00
sballe_I underfornately have a day full of meetigns tomorrow and cannot attend15:01
jed56I will take a look to give you the minimal list15:01
vincentfrancoisehvprash: do you want to discuss you BP now?15:01
pav0jed56: the hardes think will be to compare which variables are deprecated between 0.24.0 and master15:02
pav0because in 0.24.0 there is watcher.conf.sample file :P15:03
jed56You should focus on the master branch15:03
jed56watcher is still on heavy development15:03
pav0jed56: ok15:03
jed56watcher is deployed in production yet :)15:03
vincentfrancoisepav0: you can see the watcher config file for watcher in the online doc now15:03
vincentfrancoisepav0: see
jed56in my opinion you don't have to keep the compatibility with old versions15:04
pav0jed56: but puppet should remove deprecated parameters (if there are)15:05
acabotjwcroppe : Congress meeting is at  0000 UTC in #openstack-meeting tomorrow15:05
jed56if you are using the version 0.24.0  i strongly recommend to you to upgrade your version :)15:05
pav0vincentfrancoise: thanks15:05
jed56pav0: yes15:05
vincentfrancoisepav0: I would rather wait for the n-1 release to handle old parameters15:06
vincentfrancoisesince this is the target to get into the big tent15:06
vincentfrancoisethings will move quite significantly config file-wise up until this point I believe15:07
hvprashvincentfrancoise, sure. Let me first understand the DB schema.15:07
vincentfrancoisehvprash: ok fair enough :)15:07
vincentfrancoisepav0: see this bueprint for example
jed56vincentfrancoise: probably15:08
vincentfrancoisepav0: also the BP remove the [watcher_goals] section15:08
hvprashvincentfrancoise, I think now I understand the objective of this blueprint. Is there a DB relation diagram ?15:09
vincentfrancoisehvprash: actually, I don't think there is one in the doc15:11
vincentfrancoisehvprash: give me a sec, I dig in to see if I can find one15:12
hvprashnm, will go the manual route :)15:12
pav0vincentfrancoise: ok, I read it tomorrow :015:14
vincentfrancoisehvprash: you can generate a qui-and-dirty one using "pyreverse -o png watcher/db/sqlalchemy/"15:18
vincentfrancoisethough it's far from being correct15:19
*** jwcroppe has quit IRC15:19
*** jwcroppe has joined #openstack-watcher15:19
hvprashvincentfrancoise, will do. Once I understand the relation we can discuss on the specs15:26
*** dronshaw has quit IRC15:27
*** vmahe has quit IRC15:29
*** hvprash_ has joined #openstack-watcher15:36
*** sudipto has joined #openstack-watcher15:38
*** jinquan has joined #openstack-watcher15:39
*** hvprash has quit IRC15:39
*** jinquan has left #openstack-watcher15:42
*** jinquan has joined #openstack-watcher15:42
*** jinquan has left #openstack-watcher15:42
*** alexchadin has quit IRC15:43
*** jinquan has joined #openstack-watcher15:46
*** vincentfrancoise has quit IRC15:51
*** jwcroppe has quit IRC15:57
*** jwcroppe has joined #openstack-watcher15:59
edleafesballe_: sorry, missed the meeting this morning. Just FYI: The Nova midcycle is also in Hillsboro on July 19-2116:08
sballe_I thought it was earlier16:08
edleafeWill we be fighting over rooms? :)16:09
sballe_I already have a room16:09
*** jwcroppe_ has joined #openstack-watcher16:11
*** jwcroppe has quit IRC16:14
sballe_edleafe: I am more worried about the hotels16:21
*** jwcroppe has joined #openstack-watcher16:22
*** jinquan has left #openstack-watcher16:23
*** jinquan has joined #openstack-watcher16:23
*** sudipto has quit IRC16:24
*** jinquan has left #openstack-watcher16:25
*** jinquan has joined #openstack-watcher16:25
*** jwcroppe_ has quit IRC16:25
*** jinquan has left #openstack-watcher16:29
*** jinquan has joined #openstack-watcher16:29
*** jinquan has left #openstack-watcher16:29
*** bigdogstl has joined #openstack-watcher17:09
*** bigdogstl has quit IRC17:15
*** jed56 has quit IRC17:24
*** bigdogstl has joined #openstack-watcher17:35
*** bigdogstl has quit IRC17:45
*** dronshaw has joined #openstack-watcher18:29
*** bigdogstl has joined #openstack-watcher18:32
*** bigdogstl has quit IRC18:39
*** bigdogstl has joined #openstack-watcher18:43
*** bigdogstl has quit IRC18:48
*** dronshaw has quit IRC19:15
*** dronshaw has joined #openstack-watcher19:20
*** bigdogstl has joined #openstack-watcher19:28
*** bigdogstl has quit IRC19:33
*** openstackgerrit has quit IRC19:48
*** openstackgerrit has joined #openstack-watcher19:48
*** harlowja has quit IRC20:09
*** bigdogstl has joined #openstack-watcher20:22
*** bigdogstl has quit IRC20:27
*** bigdogstl has joined #openstack-watcher20:38
*** bigdogstl has quit IRC20:43
*** wootehfoot has joined #openstack-watcher21:04
*** harlowja has joined #openstack-watcher21:10
*** dronshaw has quit IRC21:12
*** hvprash_ has quit IRC21:12
*** thorst has quit IRC21:28
*** thorst has joined #openstack-watcher21:33
*** bigdogstl has joined #openstack-watcher21:34
*** thorst has quit IRC21:37
*** bigdogstl has quit IRC21:39
*** jwcroppe has quit IRC21:59
*** pav0 has quit IRC22:04
*** akijak has quit IRC22:05
*** thorst has joined #openstack-watcher22:22
*** thorst has quit IRC22:26
*** harlowja has quit IRC22:52
*** harlowja has joined #openstack-watcher22:53
*** wootehfoot has quit IRC23:13

Generated by 2.14.0 by Marius Gedminas - find it at!