16:01:07 #startmeeting oslo 16:01:08 Meeting started Mon Mar 6 16:01:07 2017 UTC and is due to finish in 60 minutes. The chair is gcb. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:09 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:01:12 The meeting name has been set to 'oslo' 16:01:16 o/ 16:01:18 o/ 16:01:23 o/ 16:01:25 courtesy ping for amotoki, amrith, bknudson, bnemec, dansmith, dhellmann, dims,dougwig, e0ne, electrocucaracha, flaper87, garyk, gcb, GheRivero,haypo, jd__, jecarey, johnsom, jungleboyj, kgiusti, kragniz,lifeless, lxsli, Nakato, ozamiatin, rbradfor, redrobot, rloo,rpodolyaka, sergmelikyan, sileht, spamaps, sreshetnyak, sreshetnyak, stevemar,therve, thinrichs, toabctl, viktors, zhiyan, zzzeek 16:01:25 o/ 16:01:25 o/ 16:01:29 o/ 16:01:30 o/ 16:01:35 o/ 16:01:45 welcome guys :-) 16:01:59 #topic Red flags for/from liaisons 16:02:03 yo 16:02:19 hi harlowja_at_home 16:02:26 hi there! 16:02:33 Nothing from Octavia 16:02:37 All is well in Cinder land. 16:02:51 o/ 16:02:59 nothing from neutron 16:03:40 nothing from ironic 16:03:59 nothing is good news for oslo lol 16:05:26 :) 16:06:16 Indeed. 16:06:24 No news is good news. 16:09:33 hmmm, think we lost gcb :) 16:09:50 or i guess that should be :( 16:09:54 Nice quiet meeting. 16:10:03 ya, i think gcb just got a ping timeout :-P 16:10:13 or that's what my IRC client says, lol 16:10:20 so i'm guessing thats why its quiet, lol 16:10:37 gcb welcome back! 16:10:38 :) 16:11:38 please help review https://review.openstack.org/#/c/312233/, seems we need this for adopting new lib 16:11:46 harlowja_at_home: networking is unstable now :( 16:11:49 Wasn't a party without him. 16:11:54 :( 16:12:20 #poll request olso weekly time 16:12:46 #link https://framadate.org/9Q15UYR19fzdqhol 16:13:12 it seems UTC+14 win 16:13:31 That site was terrible for a color blind person. :-) 16:13:50 jungleboyj, agree 16:14:22 johnsom, sorry it's not a good time for you 16:14:36 lol 16:14:47 gcb would it be useful to alternate, one week in a more asia friendly time 16:14:53 and the next week in a more friendly US time? 16:15:25 There is not much difference among colors 16:15:39 Maybe I just need to find a team member that is east coast time to liaison 16:16:17 harlowja_at_home, it's open to discuss anything :-) 16:16:39 harlowja_at_home, it's too early for you right ? 16:16:51 i'll figure it out :) 16:16:55 don't worry about me, lol 16:16:59 10 vs 8 votes is fairly close 16:17:14 but 10 vs 6 ... not so close 16:17:55 harlowja_at_home, you are still the superman as I know lol 16:18:14 gcb only a tiny superman 16:18:15 lol 16:18:48 rloo, does UTC+14 work for you ? 16:19:20 gcb: so i didn't vote cuz i don't plan on attending oslo meetings in the future. just waiting for someone else to volunteer from ironic... 16:20:19 rloo, many ironic guys from Europe , I think that works for them 16:21:02 gcb: yeah. most ironic are east coast or europe now. 16:21:20 gcb: so utc+14 should work 16:21:50 gcb: but on monday, dunno :) 16:22:37 okay, so let's try to UTC+14, and we can adjust the time as this time if someone need 16:23:47 #topic Open discussion 16:24:11 I noticed the tooz drivers policy patch got abandoned. 16:24:18 I don't really have time to chase it anyway.. 16:24:21 sooo i noticed https://github.com/harlowja/failure needs some fixup also 16:24:27 i'll try to get that done this week 16:24:30 but it came up during an arch-wg discussion at the PTG 16:24:44 hmmm, SpamapS was there an overview of that discussion? 16:24:49 (seeing as I wasn't there) 16:25:04 #link https://review.openstack.org/#/c/240645/ 16:25:07 Yeah I sent to the ML.. but basically we just want to make sure we are supportive of DLM adoption 16:25:24 110% support 16:25:24 lol 16:25:25 Especially since we will at some point discuss DLM as a base-service. 16:25:53 well here's the thing, and jd can correct me if i am wrong 16:26:36 for tooz drivers, all of them have annotations with https://github.com/openstack/tooz/blob/master/tooz/coordination.py#L41 16:27:13 so users of tooz get to choose what they care about when using the getcoordinator call @ https://github.com/openstack/tooz/blob/master/tooz/coordination.py#L720 16:27:19 so that's sort of the policy we put in place 16:27:28 if people don't give a crap about that, then they don't have to use it 16:27:44 speaking of DLM, have you seen this harlowja_at_home ? https://github.com/coreos/zetcd 16:27:49 o/ 16:27:59 dims_ haven't seen that, ha 16:28:00 funny 16:28:19 dims_ o/ 16:28:31 harlowja_at_home: OK, yeah being able to choose your guarantees at initialization time definitely makes the policy less important. 16:28:41 SpamapS so ya, there is the policy :-P 16:28:43 use it or don't, lol 16:28:58 kicking out drivers was the other idea, i'm pretty sure that would cause tooz ----> go out of openstack (via fork or other) 16:29:03 harlowja_at_home: since when we have the base-services discussion we can say "a service with a tooz driver that supports guarantees x, y, z" 16:29:14 SpamapS that's fine with me 16:29:19 Yeah, ok 16:29:34 that closes the loop for me, and I support abandoning prose for code. :) 16:29:50 ya, that was the code i came up with as a good enough compromise 16:29:59 cause kicking out drivers wasn't going to end well... 16:31:07 dims_ i wonder how fully functional that actually is 16:31:26 haven't kicked the tires harlowja_at_home. saw it float by late last week 16:31:31 ya, interesting 16:31:43 now zookeeper folks will just make a etcd wrapper api 16:31:50 and we can run this zetcd ontop of that 16:32:02 zookeeper on zookeeper 16:32:03 lol 16:32:13 lol 16:32:21 wait, is that a joke or forealz? 16:32:29 which part? 16:32:30 lol 16:32:35 zetcd compat 16:32:40 :) 16:32:41 joke 16:32:44 good 16:32:45 https://github.com/coreos/zetcd#zetcd seems real 16:32:59 but zookeeper on zetcd on etc on zookeeper is a joke 16:33:21 oh weird so there is an etcd that speaks zookeeperese 16:33:29 seems like it (or at least it tries to) 16:33:33 that seems like a recipe for destruction since zookeeper has guarantees etcd can't make 16:33:51 ya, that's what i was wondering, how compat SpamapS is that really :-P 16:34:05 compat enough to delete your whole datacenter accidentally ;) 16:34:25 cattle not pets 16:34:26 lol 16:36:20 anyone intersted in actually putting https://github.com/harlowja/failure in oslo.messaging and nova (and wherever else the same kind of code got repeated) 16:37:06 ie https://github.com/openstack/nova/blob/master/nova/cells/messaging.py#L1871-L1950 (this crap) 16:37:14 which also exists in another varation in oslo.messaging 16:37:16 and maybe more? 16:37:46 harlowja_at_home: I'd like to see it in o.m. myself... 16:38:05 ya, i need to scan all of openstack again for that kind of stuff 16:38:15 i think there are just 2 or 3 implemenations 16:38:24 depends on how many people found and copied that code,l ol 16:39:32 harlowja_at_home: perhaps add it to oslo.serialization? 16:40:09 well the library should have an API that's pretty easy to use, i think it depends on oslo.serialization already (not vice versa) 16:40:15 Not seeing it in Cinder. 16:40:24 oh nm, https://github.com/harlowja/failure/blob/master/requirements.txt#L10 i guess 16:40:25 harlowja_at_home: gotcha 16:40:59 though i do need to make https://failure.readthedocs.io/en/latest/ more useful, lol 16:41:09 jungleboyj ya, thanks for not duplicating code like nova did, lol 16:41:13 +10 :) 16:41:28 :-) Wish I could take credit. 16:41:34 harlowja_at_home: ... a documentation 'failure'??? :) Baa-ding! 16:41:43 :) 16:42:14 i need like a sexy logo for that also 16:42:33 (for the failure library) 16:42:42 (just incase people thinking other sexy logos) 16:43:05 * jungleboyj is thinking of famous wardrobe failures. 16:43:18 That or a wheel coming off a car. 16:43:38 harlowja_at_home, a logo like taskflow ? 16:44:00 i think taskflow logo i found on some free/open clipart site 16:44:00 lol 16:44:29 at least the one @ https://wiki.openstack.org/wiki/TaskFlow#Summary 16:44:39 * harlowja_at_home am not artist 16:44:39 lol 16:45:12 :-) 16:47:52 harlowja_at_home, we discussed failure in the PTG, we need more work like documentation, then we can spread it to other projects like Nova 16:47:56 ya 16:48:02 agreed 16:49:16 you just created another library, master of library creator lol 16:49:20 thanks 16:49:41 1 million 16:50:37 lol 16:50:58 so any other things ? 16:52:43 harlowja_at_home, each project team has dedicated session about project updates in Boston summit, the new library is a good one to talk 16:52:58 about 16:53:59 gcb sounds good 16:54:21 okay, I wonder we can call it a meeting now 16:54:33 Sounds good. Thank you. 16:54:41 thanks everyone 16:54:53 #endmeeting