<@nirik:matrix.scrye.com>
16:00:11
!startmeeting Infrastructure (2025-10-09)
<@meetbot:fedora.im>
16:00:12
Meeting started at 2025-10-09 16:00:11 UTC
<@meetbot:fedora.im>
16:00:13
The Meeting name is 'Infrastructure (2025-10-09)'
<@james:fedora.im>
16:00:16
Hey
<@nirik:matrix.scrye.com>
16:00:18
!info About our team: https://docs.fedoraproject.org/en-US/cle/
<@nirik:matrix.scrye.com>
16:00:18
!topic Hola y bienvenido
<@nirik:matrix.scrye.com>
16:00:18
!info Fedora Infra documentation: https://docs.fedoraproject.org/en-US/infra
<@nirik:matrix.scrye.com>
16:00:18
!info Agenda is at: https://board.net/p/fedora-infra
<@nirik:matrix.scrye.com>
16:00:18
!chair @nirik:matrix.scrye.com @zlopez:fedora.im @nb:fedora.im @dtometzki:fedora.im @jnsamyak:matrix.org @james:fedora.im
<@nirik:matrix.scrye.com>
16:00:18
!meetingname infrastructure
<@meetbot:fedora.im>
16:00:19
The Meeting Name is now infrastructure
<@nirik:matrix.scrye.com>
16:00:30
Ahoy!
<@nirik:matrix.scrye.com>
16:01:58
Might be a small crowd today... several people are out/traveling...
<@nirik:matrix.scrye.com>
16:02:00
!info Getting Started Guide: https://docs.fedoraproject.org/en-US/infra/gettingstarted/
<@nirik:matrix.scrye.com>
16:02:00
!info This is a place where people who are interested in Fedora Infrastructure can introduce themselves
<@nirik:matrix.scrye.com>
16:02:00
!topic New folks introductions
<@nirik:matrix.scrye.com>
16:02:05
anyone new around ?
<@smooge:fedora.im>
16:03:21
hi my name is Stephen Smoogen. I have not been system admining for 8 minutes now
<@nirik:matrix.scrye.com>
16:03:39
hello smooge!
<@nirik:matrix.scrye.com>
16:05:13
ok, then I guess lets move on
<@nirik:matrix.scrye.com>
16:05:16
!topic Next chair
<@nirik:matrix.scrye.com>
16:05:16
!info chair 2025-10-30 - ???
<@nirik:matrix.scrye.com>
16:05:16
!info chair 2025-10-23 - ???
<@nirik:matrix.scrye.com>
16:05:16
!info chair 2025-10-16 - zlopez
<@nirik:matrix.scrye.com>
16:05:16
!info magic eight ball says:
<@nirik:matrix.scrye.com>
16:05:26
anyone want to take the 23rd or 30th chair?
<@nirik:matrix.scrye.com>
16:05:33
or shall we just decide those next week?
<@nirik:matrix.scrye.com>
16:06:51
ok, next week it is.
<@nirik:matrix.scrye.com>
16:06:55
moving on to announcements
<@nirik:matrix.scrye.com>
16:07:00
!info CLE Infra&Releng NA-hours team has a Monday through Thursday 30 minute meeting going through tickets at 1900 UTC in https://matrix.to/#/#meeting-3:fedoraproject.org
<@nirik:matrix.scrye.com>
16:07:00
!info f43 final freeze started on oct 7th
<@nirik:matrix.scrye.com>
16:07:00
!topic announcements and information
<@nirik:matrix.scrye.com>
16:07:00
!info CLE Infra&Releng EU-hours team has a Monday through Thursday 30 minute meeting going through tickets at 0815 UTC in https://matrix.to/#/#meeting-3:fedoraproject.org
<@nirik:matrix.scrye.com>
16:07:00
!info Started working on Fedora Infra application list in hackmd https://hackmd.io/@fnF231raRruGKZbPBWKsHQ/B1F1QhRseg [zlopez]
<@nirik:matrix.scrye.com>
16:08:28
any other announcements?
<@nirik:matrix.scrye.com>
16:09:27
!topic Oncall
<@nirik:matrix.scrye.com>
16:09:27
!info on call from 2025-10-17 to 2025-10-23 - ???
<@nirik:matrix.scrye.com>
16:09:27
!info on call from 2025-10-10 to 2025-10-16 - ???
<@nirik:matrix.scrye.com>
16:09:27
!info on call from 2025-10-03 to 2025-10-09 - nirik
<@nirik:matrix.scrye.com>
16:09:27
!info https://docs.fedoraproject.org/en-US/infra/day_to_day_fedora/#_the_oncall_role_in_our_team
<@nirik:matrix.scrye.com>
16:09:45
I saw no oncall pings, I still have a todo to start a discussion on what we want to do with oncall.
<@nirik:matrix.scrye.com>
16:09:56
I can just take this next week too and try and start that discussion.
<@nirik:matrix.scrye.com>
16:10:12
so, that takes us to monitoring!
<@nirik:matrix.scrye.com>
16:10:20
!topic Monitoring discussion [nirik]
<@nirik:matrix.scrye.com>
16:10:20
!info Go over existing items and fix them
<@nirik:matrix.scrye.com>
16:10:20
!info https://nagios.fedoraproject.org/nagios
<@nirik:matrix.scrye.com>
16:11:02
basically the same as last week on the nagios side... some logdetective alerts and a cert alert this morning.
<@nirik:matrix.scrye.com>
16:11:35
we have been looking at some proxies alerting for disk space every day...
<@nirik:matrix.scrye.com>
16:11:39
it's a weird issue.
<@nirik:matrix.scrye.com>
16:13:11
zabbix alerts have been diminishing as we adjust thresholds.
<@nirik:matrix.scrye.com>
16:13:55
FYI, for anyone interested... prod zabbix alerts go to #zabbix:fedora.im and staging ones go to #noc-zabbix-testing:fedora.im
<@nirik:matrix.scrye.com>
16:14:10
(but you likely want to mute those rooms :)
<@nirik:matrix.scrye.com>
16:14:25
any monitoring questions?
<@nirik:matrix.scrye.com>
16:15:45
ok, given the low turnout today, I think perhaps we should head to open floor...
<@nirik:matrix.scrye.com>
16:15:49
!topic open floor
<@gwmngilfen:fedora.im>
16:16:58
!hi
<@zodbot:fedora.im>
16:16:59
Greg Sutcliffe (gwmngilfen) - he / him / his
<@james:fedora.im>
16:17:06
And then there was one more!
<@gwmngilfen:fedora.im>
16:17:11
sorry i'm late, family stuff. just showing my face before I go cook 😉
<@gwmngilfen:fedora.im>
16:18:06
staging alerts go to #fedora-zodbot:fedora.im actually, I think
<@nirik:matrix.scrye.com>
16:18:21
oh... right.
<@nirik:matrix.scrye.com>
16:18:29
we should standardize :)
<@gwmngilfen:fedora.im>
16:18:42
fyi I'll be doing a recording of some Zabbix show'n'tell on Monday, we'l try to get the vod out for that pronto
<@gwmngilfen:fedora.im>
16:18:56
we can maybe do a chat QnA afterwards?
<@gwmngilfen:fedora.im>
16:19:01
makes for a decent learning topic
<@nirik:matrix.scrye.com>
16:19:16
sure. sounds good!
<@gwmngilfen:fedora.im>
16:19:33
on the proxies - they *all* do the disk thing, but only some hit the alert
<@nirik:matrix.scrye.com>
16:20:15
yeah.
<@gwmngilfen:fedora.im>
16:20:19
stacked zabbix graphs ftw 😛
<@gwmngilfen:fedora.im>
16:20:34
i have no further insight, but I thought you might like to see that
<@nirik:matrix.scrye.com>
16:21:05
It's bound to be cron job triggered, but logrotate doesn't fully explain it. but we can see.
<@james:fedora.im>
16:21:09
Can you stack different days easily? Becuase the exact amount of data seems to change, but it's less obvious from collectd how much.
<@gwmngilfen:fedora.im>
16:21:19
sure
<@james:fedora.im>
16:21:24
Like maybe it's the same amount every monday, which would be interesting
<@gwmngilfen:fedora.im>
16:21:30
let me see if i can do a quick tutorial
<@gwmngilfen:fedora.im>
16:21:50
so head to zabbix.fpo > monitoring > latest data
<@james:fedora.im>
16:22:08
Wait ... I thought you were like a scottish AI assistant just giving me answers I mostly trust ;)
<@gwmngilfen:fedora.im>
16:22:23
hah
<@gwmngilfen:fedora.im>
16:22:30
then select "Proxies" as a host group in the filter
<@gwmngilfen:fedora.im>
16:22:48
and "FS [/]: Space: Used, in %" in the Name field (name of the item)
<@gwmngilfen:fedora.im>
16:23:01
if we're lucky this will work: https://zabbix.fedoraproject.org/zabbix.php?name=FS%20%5B%2F%5D%3A%20Space%3A%20Used%2C%20in%20%25&evaltype=0&tags%5B0%5D%5Btag%5D=&tags%5B0%5D%5Boperator%5D=0&tags%5B0%5D%5Bvalue%5D=&show_tags=3&tag_name_format=0&tag_priority=&state=-1&filter_name=&filter_show_counter=0&filter_custom_time=0&sort=host&sortorder=ASC&show_details=0&action=latest.view&groupids%5B%5D=38
<@gwmngilfen:fedora.im>
16:23:34
then select all and display graph - then you can adjust the timeframe of the graph
<@gwmngilfen:fedora.im>
16:23:50
fwiw it seems consistent - last 7 days:
<@james:fedora.im>
16:25:24
Yeh, psychedelic graphs ;)
<@gwmngilfen:fedora.im>
16:25:51
Zabbix is the best
<@gwmngilfen:fedora.im>
16:26:23
(i have no idea if /rainbow renders on other clients than Element, so that joke may be lost)
<@james:fedora.im>
16:27:10
Yeh, just checked and yesterday was slightly different to last Wed.
<@james:fedora.im>
16:27:24
So who knows ... back to span camping, I guess.
<@james:fedora.im>
16:27:32
So who knows ... back to spawn camping, I guess.
<@gwmngilfen:fedora.im>
16:27:35
I'm sure it would be possible to create some dependant thing in Zabbix that logs what processes are running when it alerts ... but I haven't tried that yet
<@gwmngilfen:fedora.im>
16:28:55
otherwise, yes, spurious alerts are dropping, though there is more I think we can do there. we might well want to copy the list of nagios "services" like mail: true/false etc at some point and make use of it
<@gwmngilfen:fedora.im>
16:29:37
eg right now *everything* gets the postfix queue monitor, which is less useful on some hosts - but then again, still useful with the right threshold
<@gwmngilfen:fedora.im>
16:29:46
we will iterate ofc
<@gwmngilfen:fedora.im>
16:30:28
anyway! time too cook, flying visit 😛
<@gwmngilfen:fedora.im>
16:30:33
anyway! time to cook, flying visit 😛
<@nirik:matrix.scrye.com>
16:31:51
happy dinner
<@nirik:matrix.scrye.com>
16:32:05
looking at it logrotate takes ~30min.
<@nirik:matrix.scrye.com>
16:32:21
anyhow...
<@nirik:matrix.scrye.com>
16:32:38
if nothing else, lets end here and go back to it.
<@james:fedora.im>
16:32:59
Sure, sounds good.
<@nirik:matrix.scrye.com>
16:33:04
!endmeeting