14:00:28 <tellesnobrega> #startmeeting sahara 14:00:29 <openstack> Meeting started Thu Aug 2 14:00:28 2018 UTC and is due to finish in 60 minutes. The chair is tellesnobrega. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:30 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:32 <openstack> The meeting name has been set to 'sahara' 14:01:29 <tellesnobrega> show off hands 14:01:30 <tellesnobrega> o/ 14:01:33 <jeremyfreudberg> o/ 14:02:27 <tellesnobrega> tosky? 14:04:17 <jeremyfreudberg> .. 14:04:33 <tellesnobrega> #topic News/Updates 14:04:57 <tosky> sorry 14:04:59 <tosky> here 14:05:06 <tellesnobrega> Sahara R3 was release last week as well as saharaclient 2.0 14:05:09 <jeremyfreudberg> welcome tosky 14:06:13 <tellesnobrega> yesterday I went through storyboard bugs, we might need to go through it during ptg and try to prioritize the bugs 14:06:25 <tellesnobrega> so we can try to bring those numbers down 14:07:25 <jeremyfreudberg> i'll be testing s3 on cdh soon, if there is a bug found i will try to get it included in rocky, same idea for apiv2 bugs 14:07:31 <jeremyfreudberg> beyond that i have some doc to write 14:07:33 <tosky> you can create a board or worklist for that 14:07:46 <tellesnobrega> tosky, good idea 14:08:13 <jeremyfreudberg> oh, and not sure if tellesnobrega saw my message the other day, but my tsp request was approved 14:08:21 <jeremyfreudberg> for this ptg 14:08:35 <tellesnobrega> jeremyfreudberg, did not see it. That is awesome 14:09:39 <jeremyfreudberg> tosky: you won't be there, right? 14:10:12 <tellesnobrega> I think he will make it, not sure just yet 14:10:15 <tosky> I'm probably going to be there 14:10:32 <jeremyfreudberg> oh, nice, keep us posted 14:10:56 <tellesnobrega> more updates? 14:12:06 <tellesnobrega> lets move on 14:12:20 <tellesnobrega> #topic sahara-extra 14:12:50 <tellesnobrega> SotK, about sahara-extra I talked with doug and sean yesterday and the idea is to backport everything to stable/rocky 14:13:13 <tellesnobrega> I can write the patches 14:13:19 <tellesnobrega> I plan to do it today 14:13:51 <tellesnobrega> the alternative was to drop stable/rocky branch and the work was going to be too much and involve too many people, since there are only 5 patches, we can backport easily 14:14:21 <jeremyfreudberg> backporting is fine 14:14:42 <jeremyfreudberg> i guess the bigger question is, should we adjust the release cadence of sahara-extra from cycle-with-whatever (i'm not looking at the file right now) to independent 14:15:20 <jeremyfreudberg> i think there is value in releases/tags for sahara-extra, but they don't seem to relate to the main release cycle 14:15:29 <tellesnobrega> I have to check, but the way it is right is ok, it was a mistake by the script and myself to merge the patch to create stable/rocky 14:15:33 <tellesnobrega> I shouldn't have created that 14:16:48 <jeremyfreudberg> tosky: thoughts? 14:17:01 <tosky> but which patch exactly created stable/rocky? 14:17:16 <tosky> that's the part where I didn't get - for some older M release? 14:17:34 <tellesnobrega> tosky, https://review.openstack.org/#/c/586182/ 14:17:38 <tosky> because I don't remember recent 5 patches, so it must have happened some time ago 14:17:59 <tellesnobrega> it created based on the last release 14:18:13 <tellesnobrega> basically was a timing issue, it created as a library 14:18:20 <tellesnobrega> so it picked up stuff from last release 14:18:24 <jeremyfreudberg> looks more like sean did it by accident: https://github.com/openstack/releases/commits/master/deliverables/rocky/sahara-extra.yaml 14:18:39 <jeremyfreudberg> it went from library->other, but sean missed that 14:18:48 <tellesnobrega> they have a script that does that, and it failed to detect the type 14:18:57 <tellesnobrega> exactly 14:21:05 <tellesnobrega> anything else on this? 14:21:18 <jeremyfreudberg> i think we are all on the same page on this, with at least a short term plan 14:21:24 <tellesnobrega> yes 14:21:28 <tellesnobrega> lets move on 14:21:40 <tellesnobrega> #topic Stein PTG 14:22:09 <tellesnobrega> I will start an etherpad, probably next week, so we can start planning the PTG discussions 14:22:24 <tellesnobrega> keep an eye for it and start thinking on topics for the week 14:23:42 <jeremyfreudberg> btw, i will most likely only be at ptg on sahara days, leaving on wednesday 14:24:03 <jeremyfreudberg> pending confirmation from the Foundation's travel agent 14:24:21 <tellesnobrega> sounds good, I will be the whole week, if you need me to attend something I can do it 14:25:29 <jeremyfreudberg> tellesnobrega: thanks! 14:25:32 <jeremyfreudberg> i'll let you know 14:25:36 <tellesnobrega> cool 14:26:02 <tellesnobrega> I don't have anything specific to discuss today 14:26:14 <tellesnobrega> do you have anything else? 14:26:58 <jeremyfreudberg> well, we still have the hardware for CI. we still don't really know how make sahara-ci-config work, and i've had no time 14:27:01 <jeremyfreudberg> but, just a reminder that it exists 14:27:17 <tellesnobrega> that might be better for PTG 14:27:23 <tellesnobrega> face 14:27:33 <jeremyfreudberg> yes, the actually work we can save till then 14:27:41 <jeremyfreudberg> or even if not the work, the discussion 14:27:46 <tellesnobrega> face to face we might come up with a quicker solution 14:27:47 <jeremyfreudberg> just wanted to drop the reminder now 14:27:53 <tellesnobrega> thanks 14:28:12 <jeremyfreudberg> that's it from me 14:28:21 <jeremyfreudberg> tosky: anything from you? 14:30:26 <tosky> not for now about the PTG 14:30:42 <tellesnobrega> tosky, any other topic? 14:31:39 <tosky> S3 vs S3a, but I don't have resources to try to fix it 14:32:02 <tosky> it impacts the tests I'm writing 14:32:23 <tellesnobrega> ok 14:32:27 <tellesnobrega> let me change topics 14:32:29 <jeremyfreudberg> tosky: assume that the current s3 types will stay the same. if there's a change like that it will be accompanied by a whole new interface 14:32:34 <tellesnobrega> #topic S3 vs S3a 14:34:01 <tosky> I don't think this change would require a new interface 14:35:06 <jeremyfreudberg> tosky, it wouldn't require a new interface, no 14:35:18 <tosky> it's just allowing the users to use s3:/ URI also for data sources 14:35:30 <tosky> (or the job binaries, I always mix them) 14:35:44 <jeremyfreudberg> it's the DS which uses s3a now 14:38:13 <jeremyfreudberg> tosky: is your only complaint about the abstraction the s3 vs s3a, or do you have other ideas? 14:38:34 <tosky> only the abstraction 14:39:30 <jeremyfreudberg> you mean, only the url format 14:39:35 <jeremyfreudberg> nothing else about the abstraction 14:40:50 <tosky> yes, exactly 14:42:07 <jeremyfreudberg> so, should it just allow either protocol? (and munge s3->s3a as needed?) 14:42:31 <tellesnobrega> I think we should use just one s3 14:42:37 <jeremyfreudberg> well, i guess you also have your complaint about the http/https for the endpoint 14:42:41 <tellesnobrega> and munge s3->s3a where needed 14:44:36 <tosky> that's maybe more tricky, but s3/s3a is immediately visible 14:44:58 <jeremyfreudberg> the http/https thing is actually more bothersome to me 14:45:03 <jeremyfreudberg> but, to each his own 14:46:42 <jeremyfreudberg> the http/https thing is already more solidified though, because of the --s3-ssl flag in OSC, so that's harder to backpedal 14:47:00 <jeremyfreudberg> the s3->s3a we could really just sneak in now with nobody noticing (if we supported both formats) 14:49:10 <tosky> yes, that's what I thought as well 14:49:29 <tellesnobrega> I think it works 14:51:07 <jeremyfreudberg> okay, i think i can write a patch that will allow either s3 or s3a to go in 14:51:31 <jeremyfreudberg> and change the doc, though, to say that s3 is the 'real' way 14:51:42 <jeremyfreudberg> and i think we can sneak that into rocky 14:51:56 <tosky> thanks 14:52:05 <jeremyfreudberg> rc1 is a week from today, right? 14:52:14 <tosky> yes 14:52:18 <jeremyfreudberg> excellent 14:52:20 <tellesnobrega> yes 14:52:22 <tellesnobrega> we have time 14:53:28 <jeremyfreudberg> anything else? 14:53:31 <tellesnobrega> thanks tosky for bringing it up and jeremyfreudberg for the work 14:53:34 <tellesnobrega> not for me 14:53:36 <jeremyfreudberg> np 14:53:40 <kaiokmo> tellesnobrega: can I include some quick info/update to the meeting? it's OSA related 14:53:49 <jeremyfreudberg> kaiokmo, go ahead 14:54:24 <tellesnobrega> yes please 14:55:00 <kaiokmo> ok. so, as long as rocky release is near, and ubuntu bionic is now supported on openstack, the osa team is working to add bionic support on all roles, including sahara 14:56:12 <kaiokmo> the work seems pretty consistent, and looks like sahara support for bionic is almost done, along with existing xenial, centos7 and opensuse ones 14:56:46 <tosky> thanks for that, I've seen few reviews on openstack-ansible-os_sahara 14:57:16 <tellesnobrega> kaiokmo, awesome, keep us posted and ping us if needed 14:57:32 <kaiokmo> tellesnobrega: no problem. will do 14:58:11 <tosky> kaiokmo: is there any plan to switch the Bionic jobs to python 3, or are they going to stay python 2 for now? 14:59:22 <kaiokmo> I don't know about that, but I've seen some patches related to python3 15:00:36 <openstack> mlavalle: Error: Can't start another meeting, one is in progress. Use #endmeeting first. 15:00:49 <tosky> tellesnobrega: I think you can close now 15:00:49 <kaiokmo> on xenial, at least, there's on-going work on python3. 15:00:57 <tellesnobrega> will do 15:00:57 <tosky> we are out of time and people are pushing 15:00:58 <kaiokmo> I can update you guys later on sahara channel 15:01:00 <tellesnobrega> #endmeeting