18:02:28 <phsmoura> #startmeeting Fedora Infrastructure Ops Daily Standup Meeting
18:02:28 <zodbot> Meeting started Tue Oct 11 18:02:28 2022 UTC.
18:02:28 <zodbot> This meeting is logged and archived in a public location.
18:02:28 <zodbot> The chair is phsmoura. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
18:02:28 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:02:28 <zodbot> The meeting name has been set to 'fedora_infrastructure_ops_daily_standup_meeting'
18:02:28 <phsmoura> #chair nirik mobrien aheath1992 smooge
18:02:28 <zodbot> Current chairs: aheath1992 mobrien nirik phsmoura smooge
18:02:28 <phsmoura> #meetingname fedora_infrastructure_ops_daily_standup_meeting
18:02:29 <zodbot> The meeting name has been set to 'fedora_infrastructure_ops_daily_standup_meeting'
18:02:29 <phsmoura> #info meeting is 30 minutes MAX. At the end of 30, its stops
18:02:29 <phsmoura> #info agenda is at https://board.net/p/fedora-infra-daily
18:02:35 <nirik> morning
18:02:52 <phsmoura> morning
18:03:12 <phsmoura> #info reminder: speak up if you want to work on a ticket!
18:03:12 <phsmoura> #topic Tickets needing review
18:03:21 <phsmoura> #info https://pagure.io/fedora-infrastructure/issues?status=Open&priority=1
18:03:38 <phsmoura> .ticket 10938
18:03:39 <zodbot> phsmoura: Issue #10938: Removal of 'dcantrel' account remaining things - fedora-infrastructure - Pagure.io - https://pagure.io/fedora-infrastructure/issue/10938
18:03:50 <phsmoura> low,low,ops?
18:04:04 <erolkskn> hi there
18:04:25 <erolkskn> welcome back nirik
18:04:30 <phsmoura> hi erolkskn
18:04:47 <nirik> phsmoura: +1
18:04:53 <nirik> thanks erolkskn.
18:05:44 <phsmoura> thats it for today, we dont have new releng tickets :)
18:06:02 <nirik> great!
18:06:14 <phsmoura> #topic Planning, Upcoming work and Open floor
18:06:33 <phsmoura> I have something for open floor
18:06:49 <nirik> I'm continuing to work on FMN with mkonecny... I had it processing yesterday, but with errors, mkonecny fixed those, but now it's not processing. ;(
18:07:01 <nirik> #info nirik is out next week
18:07:10 <nirik> #info go / no-go is this thursday
18:07:16 <nirik> (but it's pretty likely to be no-go)
18:07:34 <erolkskn> I'm creating a PoC work for what I have in my mind about fedbadges.
18:07:36 <nirik> I think thats all I had
18:08:32 <erolkskn> Instead of querying datanommer/datagrepper everytime an event occured we might just transform those events in to something like "metrics"
18:08:56 <erolkskn> that would be way more performant than current implementation
18:09:53 <phsmoura> nirik: about uptading firmware, some machines are neither responding http requests nor responding when using ipmitool. Also, Im having trouble to get the model of those machines, I could use beautifulsoup to get it from html but if you know a command that returns that would be easier
18:09:55 * nirik isn't sure he follows
18:10:36 <nirik> phsmoura: ok, if you can generate a list of the ones that aren't responding I can look at them, some might be expected, others might need password updated.
18:10:51 <Ebeneezer_Smooge> nirik, are you out starting on Friday or starting on Monday? I didn't remember if you were here for the full week.
18:10:57 <erolkskn> Well, you might say that I'm working on an ETL process for badges :)
18:11:00 <nirik> phsmoura: for dells I think there's a ipmitool one... delloem fru ?
18:11:04 <Ebeneezer_Smooge> sorry for jumping in
18:11:20 <nirik> I am here all week! (tip your waiters)
18:11:44 <erolkskn> :p
18:11:46 <nirik> and actually I don't head out on my trip until tuesday, but I took monday to have prep time...
18:12:51 <phsmoura> nirik: ok, will do after this meeting. delloem brings me nothing when ipmitool works :/
18:12:52 <nirik> erolkskn: so, that will need changes on the datagrepper side?
18:12:53 <Ebeneezer_Smooge> never tell people that. you will have meetings on Monday now :)
18:13:04 <erolkskn> nirik: no just the badges side
18:13:25 <nirik> I mean right now, badges gets the messages directly. it only queries datagrepper if it needs to restart to see what it missed.
18:13:51 <nirik> and... if it's using fedora-messaging, it wouldn't need to do that, as it would have a durable queue of pending messages on rabbitmq server
18:14:37 <erolkskn> Unfortunately that's not the whole case, for example take a look at "karma" badges. Every time a bodhi.update occurs, it tries to get all previous events for that user
18:14:58 <erolkskn> Actually that's how most of the "counting" stuff works on badges
18:16:38 <erolkskn> for example let's say that a bodhi.update occured now and fedbadges got a notification about it (whether with fedmsg or fedora-messaging). It queries datanommer to get all records, and counts them one by one
18:17:13 <erolkskn> by "all records" i mean with the same topic of incoming message
18:17:14 <nirik> sure. There's also cron jobs that do tons of queries as I recall.
18:17:41 <nirik> so you mean keep track of that locally so you don't need to query all the time?
18:17:49 <erolkskn> exactly
18:18:33 <erolkskn> instead of getting count by fetching records from datanommer, you know, we can just increment "something" on database and that would be it.
18:21:20 <nirik> sure, but it needs populating initially and means storing all that in a db... but yeah, might be nicer.
18:21:53 <erolkskn> Yeah that's what i'm trying to figure out, you know there are 250m+ previous events happened earlier
18:22:10 <erolkskn> we should process them first in this kind of system
18:24:11 <erolkskn> But either way, it seems like it's necessary to reprocess all messages from back, since we do know that there are a lot of badges given to wrong users.
18:25:10 <erolkskn> And a lot of not given badges, even though they are  well deserved
18:25:57 * nirik nods
18:27:17 <erolkskn> That's all i had
18:27:48 <phsmoura> is there anything else?
18:27:52 <Ebeneezer_Smooge> nope thanks phsmoura
18:28:01 <erolkskn> thanks phsmoura
18:28:13 <phsmoura> thanks all for attending :)
18:28:18 <nirik> thanks phsmoura
18:28:21 <phsmoura> #endmeeting