15:02:41 #startmeeting oslo 15:02:42 Courtesy ping for bnemec, jungleboyj, moguimar, hberaud, stephenfin, kgiusti, johnsom, e0ne, redrobot, bcafarel 15:02:42 #link https://wiki.openstack.org/wiki/Meetings/Oslo#Agenda_for_Next_Meeting 15:02:42 Meeting started Mon Nov 18 15:02:41 2019 UTC and is due to finish in 60 minutes. The chair is bnemec. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:02:43 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:02:45 The meeting name has been set to 'oslo' 15:02:53 o/ 15:02:54 o/ 15:03:01 o/ 15:03:02 Kind of here. 15:03:17 \o 15:03:26 o/ 15:03:27 o/ 15:03:32 o/ 15:03:35 jungleboyj: I'm impressed you're awake. :-) 15:03:48 :-) Check with me in 4 hours. :-) 15:05:10 Okay, good turnout today. Welcome everyone! 15:05:12 #topic Red flags for/from liaisons 15:05:37 Eyal proposed openstack/oslo.messaging master: Make sure minimum amqp is 2.5.2 https://review.opendev.org/694686 15:05:42 Nothing to report from Octavia 15:06:16 Nothing from Cinder. 15:06:28 o/ 15:07:32 Nothing much from the Oslo side. The major thing was the mask password changes in oslo.utils, but I don't think that affects anyone here. 15:08:09 Although I guess it's worth noting that those changers are making the mask_password functions in strutils case insensitive. 15:08:17 Nothing from Barbican 15:08:32 the tripleo guys seems to facing some related leak of password 15:08:37 I don't expect that to cause any problems, but if it does let use know, of course. 15:09:08 hberaud: Yeah, that's what the mask password stuff is intended to fix. 15:09:19 I suppose this is why they fixed that... 15:09:27 I need to look at the new patches for that too. 15:09:34 ack 15:10:11 Lots of very tripleo-specific keys in there, but I guess it's unavoidable. 15:10:16 and if someone can take a look to the stein and rocky backport too 15:10:33 they are welcome 15:10:39 thanks 15:10:49 I really wish Mistral didn't just log giant dicts all the time. 15:11:01 Yeah, I'll try to keep on top of the backports too. 15:11:11 ty 15:11:51 #topic Releases 15:12:03 I didn't get them done again last week. :-( 15:12:38 There were individual releases for oslo.utils proposed, but I need to make sure everything that has merged in other projects gets released too. 15:12:45 Will try to get that done today. 15:13:14 #topic Action items from last meeting 15:13:33 "hberaud to approve https://review.opendev.org/689611" 15:13:45 Done, thanks! 15:13:52 "bnemec to send email about virtual PTG" 15:14:02 Also done. More on that in a bit. 15:14:29 That was it for action items. Always like it when they're all done. :-) 15:14:40 #topic Virtual PTG 15:14:53 #link https://etherpad.openstack.org/p/oslo-shanghai-topics 15:15:19 I've tentatively scheduled this for next week at this time. 15:15:40 We'll be meeting virtually on the link in the etherpad. 15:15:48 sounds good 15:16:18 we might have more people attending that today due to the short notice 15:16:22 I'm not sure exactly how long it will take. Probably depends on how much discussion is needed. 15:16:43 Yeah, that was my thought as well. I just got one more notification from the Doodle poll this morning. 15:17:12 Speaking of which: 15:17:14 #link https://doodle.com/poll/8bqiv865ucyt8499 15:17:48 Eyal proposed openstack/oslo.messaging stable/train: Make sure minimum amqp is 2.5.2 https://review.opendev.org/694792 15:17:49 So far next week is the winner, but if that doesn't work for you let me know. 15:20:47 I also recommend trying out jitsi ahead of time. It's a less common video conferencing service, but it's open source so it's a better fit for this. 15:21:16 I will probably lurk in the background, if some liaison stuff is needed for neutron 15:21:18 If we have problems I have a bluejeans meeting we can fall back to as well. 15:21:45 bcafarel: Sounds good, thanks. 15:23:03 Okay, that's all I had about the PTG. 15:23:34 #topic Shanghai PTG Summary 15:23:46 Actually, one more thing just in case anyone missed it on the mailing list. 15:23:48 #link http://blog.nemebean.com/content/oslo-shanghai 15:24:12 I wrote up some thoughts about the event in Shanghai. 15:24:36 The good news is that there seems to be a fair amount of interest in Oslo. 15:24:54 Whether that translates into people actually contributing I don't know. :-) 15:27:42 #topic Weekly Wayward Review 15:27:52 #link https://review.opendev.org/#/c/650915/ 15:28:07 My reviewstats instance is down for some reason, but I had that one open and I know it was on the list. 15:28:25 I had left a question back in July, but it wasn't anything critical so I'm inclined to just go ahead and merge it. 15:28:33 Any objections? 15:30:17 Okay, it should be headed to the gate. 15:30:29 #topic Open discussion 15:30:35 Anything else this week? 15:31:50 As you know, I wrote a spec for nats driver of oslo.messaging. 15:32:31 https://review.opendev.org/#/c/692784/ 15:33:18 Yep, I really need to review that this week. 15:33:23 Akira72: I will take look soon 15:33:29 Me also 15:33:38 Thank you all. 15:33:41 I did a bit of investigation last week as I was writing up my PTG summary since the NATS driver came up at the event. 15:34:13 I see. 15:34:23 I found it. 15:34:56 I'll check it tomorrow. 15:35:12 I will try to get my thoughts recorded on the spec review too. 15:35:29 Yes, please. 15:36:48 Thanks for proposing it. It sounds very interesting based on what I've seen so far. 15:37:14 :D 15:38:10 Do I have to write another spec for msgpack_(de)serialize_msg()? 15:38:30 https://review.opendev.org/680628 15:38:34 #action bnemec, hberaud, and kgiusti to review NATS driver spec 15:39:20 NATS driver is using them. 15:40:05 Then I would say we don't need a separate spec. Maybe note in the commit message that the msgpack change is related to the NATS driver. 15:40:23 It doesn't look like that big of a change by itself. 15:41:18 I see. I'll add it. 15:42:05 I wonder if we should consider moving all the drivers to msgpack. If it's that much faster it could have a meaningful impact on OpenStack performance. 15:42:39 Would have to be done over multiple cycles with some compatibility code for rolling upgrades, but I think we could make it work. 15:43:02 I suppose the first thing would be to do some performance testing with msgpack swapped in for JSON. 15:43:17 Anyway, that's hypothetical future work. 15:43:22 good idea 15:44:35 In the case, we have to consider backward compatibility for migration. 15:45:09 Yeah, but I think we could do it since we have the versioned envelope for oslo.messaging. 15:45:25 We'd have to indicate whether a message is JSON or msgpack and do the right thing during the transition. 15:45:47 However, that's completely separate from the NATS stuff. 15:46:09 Right. 15:46:19 #idea consider replacing JSON serialization with msgpack 15:47:25 I'm also interested in using msgpack for (de)serialize_msg. 15:47:29 I'm also curious how msgpack will handle some of the types we special-cased in jsonutils. 15:47:38 But now I'm getting pretty far down the rabbit hole. :-) 15:48:09 We should discuss this all further on the reviews themselves. 15:48:21 So, anything else before we end the meeting for this week? 15:51:12 Sounds like we're done. Thanks for joining everyone! 15:51:15 #endmeeting