18:01:17 #startmeeting Fedora Infrastructure Ops Daily Standup Meeting 18:01:17 Meeting started Thu Jul 30 18:01:17 2020 UTC. 18:01:17 This meeting is logged and archived in a public location. 18:01:17 The chair is nirik. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:01:17 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:01:17 The meeting name has been set to 'fedora_infrastructure_ops_daily_standup_meeting' 18:01:17 #chair mboddu nirik smooge pingou 18:01:17 Current chairs: mboddu nirik pingou smooge 18:01:17 #meetingname fedora_infrastructure_ops_daily_standup_meeting 18:01:17 #info meeting is 30 minutes MAX. At the end of 30, its stops 18:01:17 The meeting name has been set to 'fedora_infrastructure_ops_daily_standup_meeting' 18:01:17 #info agenda is at https://board.net/p/fedora-infra-daily 18:01:18 #topic Tickets needing review 18:01:19 #info https://pagure.io/fedora-infrastructure/issues?status=Open&priority=1 18:01:55 pagure.issue.comment.added -- mohanboddu commented on ticket fedora-infrastructure#9181: "Decommission pdc-backend* ?" https://pagure.io/fedora-infrastructure/issue/9181#comment-668247 18:02:13 * mboddu is here and can update tickets 18:02:23 cool. lets see. 18:02:39 .ticket 7676 18:02:40 nirik: Issue #7676: Replace current mirror stats graph script with "velociraptorizer" - fedora-infrastructure - Pagure.io - https://pagure.io/fedora-infrastructure/issue/7676 18:03:17 was hoping smooge might be able to comment here... I am not sure what the status is of all the stats stuff... should we keep it with the current tickets, just one, more? 18:04:10 I will look at it tomorrow. They aren't equivalent 18:04:12 Why are we looking at the old ticket? 18:04:30 because I moved it to needs-review so we could see if it's still needed? 18:04:33 because it is an important thing for NEST 18:04:41 Ah okay 18:05:04 and that too 18:05:35 so, I guess we need it, so move back to waiting on asignee? or just leave it and discuss tomorrow? 18:06:11 pagure.issue.edit -- smooge edited the close_status and status fields of ticket fedora-infrastructure#7676 https://pagure.io/fedora-infrastructure/issue/7676 18:06:12 pagure.issue.comment.added -- smooge commented on ticket fedora-infrastructure#7676: "Replace current mirror stats graph script with "velociraptorizer"" https://pagure.io/fedora-infrastructure/issue/7676#comment-668249 18:06:27 I am not sure, is there anything to discuss on this? If not, change it to waiting on assignee 18:06:29 I am closing it for the newer tickets 18:06:53 wils countme and the data-analysis proxy will cover this I believe 18:07:11 I intend to fix those tomorrow 18:07:23 ok, cool. sorry to bug you on it smooge 18:07:41 its ok matt needed to head to a meeting so we are taking a break until that is done 18:07:49 .ticket 9181 18:07:50 nirik: Issue #9181: Decommission pdc-backend* ? - fedora-infrastructure - Pagure.io - https://pagure.io/fedora-infrastructure/issue/9181 18:08:21 I wonder if we are using it only for pdc-updater tho? 18:08:30 mboddu: could you use compose-x86-01 for those things? or pdc-web01? 18:08:51 yeah, I guess we need to make sure everything is moved off them, then I am 100% fine with turning them off. 18:08:59 pagure.issue.comment.added -- smooge commented on ticket fedora-infrastructure#9066: "MBS is unreasonably slow and never completes module builds" https://pagure.io/fedora-infrastructure/issue/9066#comment-668251 18:09:19 I could use any of those boxes, but we need to verify before decomissioning it 18:10:38 so med trouble / med gain / groomed? 18:11:01 +1 18:11:02 and someone can check everything there, then we turn them off (but keep them around) and try and use another host next time we need something, if it fails we bring those back? 18:11:17 Agreed 18:11:54 pagure.issue.tag.added -- mohanboddu tagged ticket fedora-infrastructure#9181: groomed, medium-gain, and medium-trouble https://pagure.io/fedora-infrastructure/issue/9181 18:11:55 pagure.issue.edit -- mohanboddu edited the priority fields of ticket fedora-infrastructure#9181 https://pagure.io/fedora-infrastructure/issue/9181 18:11:56 pagure.issue.comment.added -- mohanboddu commented on ticket fedora-infrastructure#9181: "Decommission pdc-backend* ?" https://pagure.io/fedora-infrastructure/issue/9181#comment-668252 18:11:57 pagure.issue.comment.added -- jkaluza commented on ticket fedora-infrastructure#9066: "MBS is unreasonably slow and never completes module builds" https://pagure.io/fedora-infrastructure/issue/9066#comment-668253 18:12:21 thats all needs reviews. 18:12:22 FYI, none from releng as they were groomed in the early morning session 18:12:24 anything on releng side? 18:12:29 ah, cool. 18:12:34 I have a question from network engineering 18:13:02 fire away 18:13:04 And the blocker is fixed by nirik/smooge and Matt (bz not responding to the bodhi pushes) 18:13:07 or is that... transmit when ready. 18:13:44 we have a bunch of builders trying to talk out to the internet 18:14:16 well, they do need to talk to pagure.io and a few things... 18:14:17 the hosts they seem to want to talk to is ci.centos.org on port 9940 18:14:31 huh. 18:14:52 this is generically all the builders? 18:15:38 so it comes from the build network so matt thought they were builders 18:15:44 I am doing the reverse looksup now 18:15:49 and 18:15:56 they are mbs-backend and pdc-backend 18:16:01 thats busgateway... 18:16:13 well, that ci bus gateway 18:17:19 but I am not sure why they would be sending there... 18:17:38 how about we open a ticket with the details and perhaps pingou could sort if it's needed? 18:18:44 ok so it is pdc-backend1/1/2 , mbs-backend and bodhi-backend 18:20:07 I will open a ticket 18:20:18 I think thats wrong and not needed anymore, but I am confused on how all this interacts anymore 18:20:50 they also talk to anitya-frontend 18:21:02 pdc-backend we may be able to kill 18:21:13 (if all they did was run pdc-updater, we're good) 18:21:24 well, bodhi is all on fedora-messaging no? 18:21:27 yup 18:21:36 so, it shouldn't need to do this. 18:21:51 and what is being sent/listened there? isn't centos-ci also on fedora-messaging now? 18:22:04 should yes 18:22:12 isn't fedora-messaging 9940 ? 18:22:21 could we have left over config? 18:22:27 smooge: no 18:22:34 pingou: yes 18:22:50 lol 18:22:56 roles/fedmsg/base/templates/endpoints-cico.py.j2 likely is telling things to talk to the centos-ci gateway 18:23:11 I thought fedmsg was 9940 but finding out it was 3999 also confused me 18:23:24 roles/openshift-apps/monitor-gating/templates/monitor_gating.cfg seems to be trying to talk to the resultsdb-cc-ci that no longer exists 18:23:52 fedmsg is many ports. 9940 is the bus gateway... ie, you want to connect to another bus and listen/push messages... 18:24:26 3999 is a relay... so connect to it and it sends on to another machine(s) 18:24:45 3000-3099 is machines talking to each other 18:25:16 fedora-messaging is 5671 or 15671 18:25:29 I'm going to ask the fedora-ci folks for confirmation for fedora-messaging 18:25:39 if that's the case, we may be able to clean up some route/config 18:25:53 we really need to figure out whats left on fedmsg and make sure it's only talking to each other/busgateways. 18:26:18 anyhow, a ticket might be good to sort this. 18:26:25 anything else before we close? 18:26:41 nope 18:26:54 Nothing from me 18:27:39 thanks for coming everyone 18:27:42 #endmeeting