15:01:59 #startmeeting climate 15:02:00 Meeting started Fri Apr 4 15:01:59 2014 UTC and is due to finish in 60 minutes. The chair is DinaBelova. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:02:01 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:02:03 The meeting name has been set to 'climate' 15:02:08 o/ 15:02:13 hello folks! 15:02:13 Hi 15:02:13 \o 15:02:18 hello 15:02:21 o/ 15:02:26 hello, guys) 15:02:44 ok 15:02:53 so today's agenda is the following 15:03:00 #link https://wiki.openstack.org/wiki/Meetings/Climate 15:03:11 #topic Action items from the last meeting + reviews queries 15:03:20 o/ 15:03:22 #link http://eavesdrop.openstack.org/meetings/climate/2014/climate.2014-03-28-15.00.html 15:03:25 Swann, o/ 15:03:38 so, as for the action items 15:03:52 bauzas_mob, are you ok to give status? 15:03:54 I got one :-) 15:04:20 well) 15:04:20 so the etherpad is created 15:04:40 feel free to comment it 15:04:43 ok, cool 15:04:57 may you publish the link& 15:04:59 here? 15:05:12 So I'll create action item 15:05:24 Sure 15:05:48 ok, waiting for that) 15:06:04 https://etherpad.openstack.org/p/climate-resource-allocation-framework 15:06:33 #action everyone take a look on https://etherpad.openstack.org/p/climate-resource-allocation-framework 15:06:41 schizophrenia, bauzas_mob is my mobile device :) 15:06:50 we got it) 15:06:57 two bauzas at the price of one :) 15:07:02 ok, as for the my action items 15:07:24 so Barbican folks seem to need maintenance) 15:07:33 so we can join them, I guess) 15:07:35 soo 15:08:01 #action DinaBelova write Barbican folks about one gerrit maintenance 15:08:02 cool :) 15:08:07 ok 15:08:22 as for the https://etherpad.openstack.org/p/climate-bp-basic-volume-plugin 15:08:33 I'm really sorry, but I ha dno time for it this week 15:08:55 So I'll take a look on it this one 15:09:13 #action DinaBelova view & comment https://etherpad.openstack.org/p/climate-bp-basic-volume-plugin 15:09:17 The same is for me 15:09:25 ok 15:09:52 and pafuent added https://review.openstack.org/#/c/84158/ on review) 15:09:56 that's quite cool 15:10:08 but as you see there is error now there 15:10:17 so pafuent please rebase this change 15:10:18 yup 15:10:24 OK 15:10:29 I'll do it 15:10:29 it looks like fast-forward merging did not work 15:10:31 ok 15:10:43 ok, so that's it for this topic 15:10:53 #topic Renaming status 15:11:12 I'm still waiting the response from the foundation... 15:11:18 I'll ping them one more time today 15:11:27 to find out results of their checks 15:11:40 ok 15:11:53 cool 15:11:54 as if we'll be together with barbican, we need to know what name to use :D 15:12:05 :-D 15:12:11 well, they were silent, will ping them one more time 15:12:26 #action DinaBelova ping foundation about new name search status 15:12:30 ok 15:12:36 so, I'm moving further 15:12:42 #topic Cross-project session status 15:12:49 ok, so we have some news 15:13:16 #link http://summit.openstack.org/cfp/details/45 15:13:22 first of all, it was decided to set deadline for the cross-project sessions and 'other' sessions to the 10 April 15:13:24 so 15:13:30 (for convenience :) ) 15:13:35 it'll be the last day for these topics proposal 15:13:37 ;) 15:14:09 and also it was decided to make decisions about what sessions to be spoken before mid April 15:14:27 so we'll know for sure about if there will it be ~15 April 15:14:40 currently chances are quite nice, btw 15:14:53 cool 15:15:02 I raised this topic during last Gantt meeting 15:15:02 so, that were some kind of news I wanted to tell on this topic 15:15:08 yep, saw that) 15:15:13 because there was some interest 15:15:17 lurker ! ;) 15:15:28 about gantt api? 15:15:39 this one? 15:15:40 nope, about cross-service scheduling 15:15:43 oh, ok 15:15:48 cool 15:15:56 so, I think we might move on 15:16:11 +1 15:16:16 #topic Energy efficiency BP status 15:16:29 f_rossigneux, you here? 15:16:46 yes 15:16:49 ok, cool 15:16:57 please describe current status for this BP 15:17:02 in several words 15:17:56 * DinaBelova waiting :) 15:18:12 This blueprint can be reviewed. 15:18:32 well, I was mostly about what had you already done 15:19:05 because, for example, for me it was quite a surprise to anderstand that you have started your work - without setting BP to the started phase 15:19:15 and to learn that you created separated repo, etc. 15:19:35 because I've looked through your etherpad 15:19:42 #link https://etherpad.openstack.org/p/climate-standby-modes 15:19:54 and I've got some questions 15:20:05 f_rossigneux: did you take a look at the ironic api ? 15:20:10 #link http://docs.openstack.org/developer/ironic/webapi/v1.html#nodestates 15:20:44 Not yet, today I was finalizing the client to query my "kwstandby" module. 15:20:46 if it's can be leveraged ... ? 15:21:31 ok 15:21:44 I have writte a module to put the host in standby modes. It in finished but not yet tested on a real infrastructure. Climate will use it, but I consider it as an external project. It could be merged in the future. 15:22:33 I would vote to put your plugin in the contrib repo then 15:22:34 I've got mostly questions about your events - for example, there is already pre-start events, etc. + I'm not sure that it 1/ should be separated repo 2/ it definately should not be on your github, but in the stackfoge, even if needed 3/ i definately don't like the idea of 3-rd party dependency for climate 15:22:43 The most important feature to implement now is the reservation aggregation, in order to minimize the number of active hosts. 15:22:48 so ++ for bauzas 15:23:40 because simply we'll have no opportunity to say "import kwstandbyclient" in climate 15:23:50 +1 15:23:55 as we're still hoping for being separated project after summit 15:24:12 and there *can't* be such dependencies in incubated project, for instance 15:24:23 +1 ^^^ 15:24:39 I'll ask all of us to take a look on Francois' etherpad and to comment all unliked moments 15:24:39 we can imagine a plugin in contrib with appropriate instructions for adding it 15:25:12 f_rossigneux, sorry for that, but it would have been quite easier if we *knew* you started work 15:25:15 so 15:25:22 the bad is that we won't have the opportunity to leverage it within Climate directly :( 15:25:51 #action all comment https://etherpad.openstack.org/p/climate-standby-modes - find all naughty things and try to do that as quick as possible 15:26:00 ok 15:26:16 so, if that's it here, I'll set next topic 15:26:27 ok 15:26:31 #topic Open discussion 15:26:47 so, folks, do you have some questions to be discussed here? 15:26:55 https://bugs.launchpad.net/climate/+bug/1301294 15:26:57 Launchpad bug 1301294 in climate "Physical reservations don't work when service name != hypervisor_name" [Undecided,New] 15:27:21 ok, looking on it 15:27:29 bauzas: I don't get why are you saying "Nova puts the host in the freepool agg and take use of service_name." Climate is passing the id or the name of the host to nova. 15:27:31 sure pafuent 15:27:45 so, let me show you the discrepancy 15:27:52 OK 15:27:54 that's happening with Havana 15:28:07 but there are possibilities to get it in also Icehouse 15:28:49 https://github.com/stackforge/climate/blob/master/climate/plugins/oshosts/reservation_pool.py#L194 15:29:02 agg.hosts is returning the name of the hosts 15:29:29 but the names are actually the service_name's :) 15:29:34 not the hypervisor ones 15:29:40 that's highly visible in Havana 15:29:44 wow, nice) 15:29:47 because on devstack 15:29:58 hypervisor_hostname is the FQDN 15:30:11 like foo.local 15:30:19 and service_name is only foo 15:30:31 test it on your own and you'll see 15:30:37 sorry.. What's FQDN? 15:30:45 fully qualified domain name 15:30:53 ah OK! 15:30:58 http://en.wikipedia.org/wiki/Fully_qualified_domain_name 15:31:00 ;) 15:31:07 got it, thanks! 15:31:12 and to cover all cases, the hostname can be set in nova.conf too 15:31:18 yep 15:31:21 ++ 15:31:25 hence the problem 15:31:28 so, ok 15:31:35 the service_name will be set by nova.conf 15:31:39 pafuent, have you got the problem? 15:31:45 and possibly different from hypervisor_hosntame 15:32:10 so we need to rely on service_name 15:32:14 for the test 15:32:31 pafuent? 15:32:48 meaning that the arg 'host' must be coming from host['service_name'] 15:32:59 and not host['hypervisor_hostname'] 15:33:01 well, I guess that's it here 15:33:07 some other things? 15:33:12 so I was proposing to store it in DB 15:33:25 because we already obtain it when inventoring the host 15:33:35 * DinaBelova feeling bad because of catching a cold, please let's end this meeting :) 15:33:36 that's only a value we don't take 15:33:36 I'm a little lost, I need to read this carefully 15:34:20 folks, I'm going to die right now, let's continue this discussion in our channel 15:34:21 just a quick work about notificaiton ... 15:34:23 ok 15:34:26 scroiset_? 15:34:27 I'reviewed it 15:34:28 https://github.com/stackforge/climate/blob/master/climate/plugins/oshosts/reservation_pool.py#L194 15:34:48 I have to go in 5 mins anyway 15:34:51 need a rebase and some words to document it but seems good :) 15:35:05 DinaBelova: if you feel bad, give me the chair 15:35:13 and go sleeping :-) 15:35:24 well, several minutes I'll be ok) 15:35:25 your health is much more important than our meeting :) 15:35:35 +1 15:35:36 agree 15:35:39 +1 15:35:49 well, as for the notifications, let's rebase it and review :) 15:36:07 and as for the pafuent 's bug -> continue in our channel 15:36:16 everyone ok with it? 15:36:19 Yes 15:36:21 yes 15:36:31 cool, thanks) 15:36:32 bye! 15:36:37 #endmeeting