15:39:50 #startmeeting Pulp Triage 2020-11-03 15:39:50 !start 15:39:50 #info fao89_ has joined triage 15:39:50 Meeting started Tue Nov 3 15:39:50 2020 UTC. The chair is fao89_. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:39:50 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:39:50 The meeting name has been set to 'pulp_triage_2020-11-03' 15:39:50 fao89_: fao89_ has joined triage 15:39:53 #info ggainey has joined triage 15:39:53 !here 15:39:53 ggainey: ggainey has joined triage 15:39:56 #info ppicka has joined triage 15:39:56 !here 15:39:56 ppicka: ppicka has joined triage 15:40:01 #info ttereshc has joined triage 15:40:01 !here 15:40:01 ttereshc: ttereshc has joined triage 15:40:05 https://hackmd.io/@pulp/triage/edit 15:40:13 #info daviddavis has joined triage 15:40:13 !here 15:40:14 daviddavis: daviddavis has joined triage 15:40:23 topic: Travis pricing - https://blog.travis-ci.com/2020-11-02-travis-ci-new-billing 15:40:27 #info dkliban has joined triage 15:40:27 !here 15:40:27 dkliban: dkliban has joined triage 15:40:33 yea so this could potentially impact us badly 15:40:36 we just had a meeting for that 15:40:38 yea 15:40:54 we agreed to do more research and wait and see 15:41:20 the backup plan is moving to github actions, the downside would be: we can't re-run single jobs 15:41:22 do we have any idea how many minutes the pulpproj org is using on travis today? 15:41:32 the whole org? 15:41:35 yeah 15:41:36 over 1000 probably 15:41:57 well yeah - but "what do we do" is different, if it's 1010, than if it's 100,000 :) 15:42:08 45,101 mins 15:42:10 per month 15:42:14 oof, ok 15:42:38 https://i.pinimg.com/originals/ae/38/6c/ae386c11446f208a85e304f0724cba48.jpg 15:42:47 and is the 1K/month per-org or per-repo, do we know? 15:42:55 dalley++ 15:42:55 fao89_: dalley's karma is now 293 15:42:56 in what way fao89 ? 15:42:58 it's unclear 15:43:03 fun 15:43:35 should we email their support and ask for their advice or at least clarification 15:43:55 mcorr: I'm not seeing pulpbot messages again 15:44:00 I think so. I think we should pre-emptively ask to raise our limit to at least 5000 15:44:02 #info x9c4 has joined triage 15:44:02 !here 15:44:02 x9c4: x9c4 has joined triage 15:44:02 Damn. And it was working the last time :( 15:44:19 otherwise we will run out very quickly 15:44:33 5000 is well short of the 45000 we use each month 15:44:37 I mean, at a minimum we need to be able to answer the question "How much is this going to cost us at current consumption?" when asked 15:44:39 yah 15:44:46 daviddavis, is that really how many we use 15:44:49 wow 15:45:04 this is based on https://travis-ci.com/github/pulp?tab=insights not sure how accurate it is tho 15:45:05 if we're using 45K/mo, 5K means we run out in a week or so 15:45:30 I'll email travis and ask some questions 15:45:35 kk 15:45:40 let's create 10 pulp orgs, hahaha 15:45:43 lol 15:45:48 each plugin has their own pulp org 15:45:53 yep! 15:45:56 yeah, I expect that would work for about a month :) 15:46:56 can we move to the next topic? 15:47:11 did y'all come to any conclusion about how much time/resource it would take to switch us to github actions? 15:47:55 ggainey, not really, although since part of the work has already been done, it would be an easier migration than gitlab CI most likely 15:47:56 I think about one day 15:48:17 one day seems really optimistic 15:48:31 we already have pulp_npm running on github actions, it is a matter of moving it to plugin-template and apply to other plugins 15:48:34 yeah ... i think a week. we have meetings and such 15:48:37 ok - I mean, if it's less than a person-week, feels like maybe next sprint we just bite the bullet, push off someting esle, and make the switch 15:48:43 what about the release stuff? 15:48:47 setting vars, etc 15:48:51 non-plugin repos 15:49:18 daviddavis: if we can get the bulk of the travis-use off of travis as a first cut, it's prob worth the exercise 15:49:30 I agree 15:50:28 anyway - sounds like The CI Gang is on top of this 15:50:34 * ggainey cheers wildly from the sidelines 15:50:37 I think it would be a good idea to ask internally what are people's experiences with Gitlab and Github Actions 15:50:42 +1 15:50:55 +1 15:51:02 +1 15:51:54 which comment are we +1-ing 15:52:06 your comment 15:52:06 "ask internally" I think 15:52:09 yus 15:52:12 yours 15:52:20 ok lol 15:52:24 just checking 15:53:13 I'm good to move on to next topic, fwiw 15:53:17 +1 15:53:28 fwiw, even if we see Gitlab as a better long-term option, if it really won't take long to move to GHA then we can do so temporarily 15:53:45 +1 to move to the next topic 15:53:48 I'm good w/stepwise-refinement 15:53:50 topic: Makefile https://github.com/pulp/plugin_template/pull/278 15:54:29 so I wanted to get some more buy in here before I committed more time to this 15:54:43 I think it's a really good idea 15:54:48 +100 15:54:51 daviddavis++ 15:54:51 ggainey: daviddavis's karma is now 379 15:55:03 cool, alright, any objections? dkliban bmbouter ttereshc 15:55:07 daviddavis: can I send you beer as a reward? :) 15:55:10 lol 15:55:12 daviddavis++ 15:55:12 fao89_: daviddavis's karma is now 380 15:55:17 I will always accept beer 15:55:17 daviddavis: no objections from me 15:55:26 no objections, big +1 15:55:36 cool 15:55:41 +1 15:55:48 just need to find the time for this then 15:55:49 +1 15:55:56 this would be a great thing 15:56:03 strongly concur 15:56:21 matrix is back! 15:56:27 would also cut down on our travbis-minutes, by letting us reliably run tests locally before pushing a PR :) 15:56:35 tru 15:57:03 that's all I had for this then 15:57:23 topic: Are there any builds we can disable in Travis? 15:57:41 I saw this run last night https://travis-ci.com/github/pulp/pulplift/builds/197596891 15:57:50 can it be disabled? 15:57:54 yes 15:58:02 * daviddavis disables it 15:58:05 anything else? 15:58:13 that's it 15:58:27 !next 15:58:28 fao89_: 2 issues left to triage: 7766, 7760 15:58:28 #topic https://pulp.plan.io/issues/7766 15:58:29 RM 7766 - dkliban@redhat.com - NEW - docs for older versions of pulpcore overwrite latest docs 15:58:30 https://pulp.plan.io/issues/7766 15:58:45 #idea Proposed for #7766: Leave the issue as-is, accepting its current state. 15:58:45 !propose accept 15:58:45 fao89_: Proposed for #7766: Leave the issue as-is, accepting its current state. 15:58:57 +1 15:59:00 +1 15:59:01 +1 15:59:12 +1 15:59:13 #agreed Leave the issue as-is, accepting its current state. 15:59:13 !accept 15:59:13 fao89_: Current proposal accepted: Leave the issue as-is, accepting its current state. 15:59:14 #topic https://pulp.plan.io/issues/7760 15:59:14 fao89_: 1 issues left to triage: 7760 15:59:15 RM 7760 - lboclboc - POST - Multiple errors in documentation 15:59:16 https://pulp.plan.io/issues/7760 15:59:23 fao89_, daviddavis if the installer / nested virt tests run on every PR, we could decrease them to nightly and/or every release 16:00:11 dalley: yea, I know nothing about the installer CI but that makes sense to me 16:00:16 for 7760, accept 16:00:16 7760 - already in post, accept and add? 16:00:20 +1 16:00:24 +1 16:00:28 +1 16:00:35 #idea Proposed for #7760: accept and add 16:00:35 !propose other accept and add 16:00:35 fao89_: Proposed for #7760: accept and add 16:00:37 #agreed accept and add 16:00:37 !accept 16:00:37 fao89_: Current proposal accepted: accept and add 16:00:38 fao89_: No issues to triage. 16:00:42 woot 16:00:49 dalley, they excercise codepaths we do not check by our molecule tests. 16:00:49 dalley, ggainey, ipanova, dkliban, let's start the meeting at :10 mins ? 16:01:00 ttereshc: works4me 16:01:08 ttereshc: yeah 16:01:09 I have to get some tea anyway 16:01:27 #endmeeting 16:01:27 !end