08:30:39 <siddharthvipul> #startmeeting CPE sustaining team stand up - Jul 20, 2020
08:30:39 <centbot> Meeting started Mon Jul 20 08:30:39 2020 UTC.  The chair is siddharthvipul. Information about MeetBot at http://wiki.debian.org/MeetBot.
08:30:39 <centbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
08:30:47 <siddharthvipul> #chair pingou jednorozec Saffronique|w mkonecny
08:30:47 <centbot> Current chairs: Saffronique|w jednorozec mkonecny pingou siddharthvipul
08:31:08 <jednorozec> hello
08:31:14 <siddharthvipul> hope I didn't miss anyone (will write agenda pad today)
08:31:18 <siddharthvipul> jednorozec: pingou \o "D
08:31:23 <siddharthvipul> mkonecny: helloww
08:31:33 <siddharthvipul> one again, sorry for false alarm ~40 minutes early
08:32:05 <siddharthvipul> I had similar "CI ticket triage" thing on my calendar that I was using for myself, got confused and didn't think it through
08:32:56 <pingou> how is everyone today?
08:33:24 <siddharthvipul> pingou: pretty good, how are you doing?
08:33:49 <pingou> doing fine :)
08:33:49 * jednorozec has a new pet
08:33:55 <pingou> congrats jednorozec !
08:34:25 * mkonecny is here
08:34:56 <mkonecny> jednorozec: What pet?
08:35:07 <jednorozec> let me upload a picture
08:35:46 <siddharthvipul> should we jump to CentOS Infra tickets? (after the picture of course) ^^
08:36:04 <pingou> siddharthvipul: +1
08:36:06 <mkonecny> I saw only two of them on the tracker
08:36:14 <pingou> priorities are priorities, we want the pictures first :D
08:36:26 <jednorozec> https://fedorapeople.org/~humaton/srnka.jpg
08:36:32 <siddharthvipul> mkonecny: I just added another one.. it's a feature request.. but we can discuss
08:36:40 <pingou> jednorozec: nice!
08:36:56 <pingou> jednorozec: do you bring them inside at night?
08:37:07 <jednorozec> she and her calves are coming to feed on cherry tree
08:37:10 <siddharthvipul> nice rabbit (you can't hit me over IRC)
08:37:17 <mkonecny> jednorozec: I always wanted one, but I will probably get more sheep :-)
08:37:49 <pingou> jednorozec: no cherry jam for you this year!
08:37:53 <jednorozec> They are hanging aroud sleeping on my meadow and eating cherry's
08:38:04 <pingou> that's the life...
08:38:21 <siddharthvipul> ^^
08:38:27 <siddharthvipul> #link https://pagure.io/centos-infra/issue/2
08:38:28 <Saffronique|w> Very cute!
08:38:28 <mkonecny> I have three of this breed planned for next year https://en.wikipedia.org/wiki/Cameroon_sheep
08:38:48 <siddharthvipul> so this is already evaluated and is waiting on bstinson , let's skip this
08:39:00 <siddharthvipul> #link https://pagure.io/centos-infra/issue/3
08:39:03 <pingou> mkonecny: looking nice :)
08:39:38 <mkonecny> pingou: I wanted a smaller breed than I have right now
08:39:56 <pingou> siddharthvipul: do you have the access to update zabbix?
08:40:50 <siddharthvipul> pingou: I am not sure if I have access because I don't know where they are.. running the infra playbook setup on those nodes should configure them all. I can take this and dig to find more
08:41:11 <siddharthvipul> just to confirm, I do have access to those nodes where zabbix nodes will be installed
08:41:26 <siddharthvipul> I might not have access to where zabbix controller is running
08:41:51 <pingou> cool
08:41:52 <Saffronique|w> Another idea, might be to use the Openshift monitoring stack instead
08:41:58 <siddharthvipul> Saffronique|w: what do you think.. we also have monitoring configuration of openshift, should there be a separate ticket?
08:42:36 <pingou> would this mean having two monitoring stacks?
08:43:13 <Saffronique|w> Hmm it would mean, keeping centosci resources being monitored by the openshift stack, and centos infra monitored by the zabbix stack
08:43:23 <Saffronique|w> they are separated by access anyway
08:43:42 <pingou> isn't one a subset of the other?
08:43:43 <siddharthvipul> pingou: hmm, I thought we were to plug openshift prometheus data in zabbix itself..  but we have access issues on zabbix so have a separate monitoring for openshift
08:43:48 <Saffronique|w> one would think :p
08:44:02 <Saffronique|w> but in practice, anything centosci is completely separate from centos infra
08:44:12 * siddharthvipul nods
08:45:06 <Saffronique|w> Going forward, probably better to move to openshift monitoring stack anyway
08:45:17 <siddharthvipul> let's discuss the details somewhere else.. Saffronique|w since you have idea of what we have to do, do you think the tags make sense?
08:45:48 <Saffronique|w> as more and more of our workloads are containerised, or kubevirtitised, and can be managed via openshfit/kubernetes, they will be easy to integrate with the openshift monitoring stack
08:45:52 <siddharthvipul> openshift monitoring is blocked on the previous ticket as well (alias for emails)
08:46:00 <siddharthvipul> Saffronique|w: I agree
08:47:33 <pingou> your call :)
08:47:53 <siddharthvipul> okay, last ticket from CentOS Infra side
08:47:57 <siddharthvipul> #link https://pagure.io/centos-infra/issue/4
08:48:00 <Saffronique|w> yeah it comes down to, we have access to one system, the openshift stack, and we dont have access to the other zabbix so
08:48:03 <Saffronique|w> vOv
08:48:19 <pingou> Saffronique|w: short term vs long term
08:49:40 <Saffronique|w> from my point of view anyway, I'd be much more in favour of the OS monitoring stack ;D
08:49:56 <Saffronique|w> Its a fight I wanna take into the Fedora infra too!
08:51:18 * pingou doesn't like fights
08:51:49 <siddharthvipul> let's jump to next ticket :P
08:52:00 <pingou> can this stack monitor things that aren't in openshift?
08:52:15 <Saffronique|w> Yep
08:52:21 <pingou> cool
08:52:32 <Saffronique|w> Similarly to zabbix where nodes run an agent
08:52:39 <Saffronique|w> you can have prometheus exporters
08:53:09 * pingou needs to create the initiative doc for monitoring in the fedora infra
08:53:36 <pingou> for #4
08:54:13 <pingou> do you need help w/ this or do you have it?
08:54:24 <siddharthvipul> so the need would be to write ansible playbook to spin a vm with requested resources and serve back.. we will have to explain the module
08:54:38 <siddharthvipul> explore*
08:54:45 <siddharthvipul> so not a lot of difficulty I will say
08:55:04 <siddharthvipul> just a little bit of research, create a duffy account from admin account and give access
08:55:08 <pingou> medium trouble medium gain?
08:56:06 <Saffronique|w> quick question related to this one, since we have kubevirt installed in the openshift cluster, users can just create a VM directly there
08:56:41 <Saffronique|w> in the longerun, maybe opennebula might not even be needed? could repurpose its nodes as openshfit nodes instead
08:56:47 <siddharthvipul> Saffronique|w: that's a nice point.. can they request custom images?
08:57:11 <Saffronique|w> not entirely sure.. we should play around with it to understand it more
08:57:24 <Saffronique|w> but it seems to be able ot create VMs using the kubernetes api..
08:58:17 <siddharthvipul> this needs proper planning.. we kinda don't want to care for other distros than fedora and CentOS
08:58:34 <siddharthvipul> but we want to make sure that the vms we offer are in good condition.. I can trust openshift for that
08:58:52 <Saffronique|w> yepyep, something to spike, outside #4
08:59:04 <siddharthvipul> Saffronique|w: fair
08:59:15 <siddharthvipul> right! so we took all our time discussing tickets
08:59:53 <siddharthvipul> pingou: is this format fine or we should try to match it as fedora-infra stand up
08:59:57 <pingou> on the Fedora-infra one there is one ticket
08:59:59 <pingou> https://pagure.io/fedora-infrastructure/issue/9149
09:00:02 <pingou> which is about the registry
09:00:03 <siddharthvipul> where we just groom/evaluate the tickets
09:00:20 <pingou> siddharthvipul: isn't that something that you looked at with Clément at one point?
09:01:07 <siddharthvipul> pingou: no, I looked at pushing container images to dockerhub.. registry and quay were out of scope because releng access
09:01:13 <pingou> okido
09:01:29 <pingou> that's the only new ticket from this week-end
09:01:35 <pingou> jednorozec: anything on your side?
09:01:58 <siddharthvipul> we did discuss that at some point we can try to replace the bash script for it in ansible so that rbac could solve the problem of access
09:02:01 <jednorozec> pingou, not musch there are 2 new unretiremtn tickets
09:02:38 <pingou> jednorozec: we really should see if we can automate some parts of that process
09:02:45 <jednorozec> and I would like to play with toddlers to se if I can unblock packages in koji based on relevant comment on unretiremnt ticket
09:02:57 <pingou> +1
09:03:07 <pingou> jednorozec: let me know if you need help with toddlers
09:03:17 * pingou still porting pdc-updater to it
09:04:38 <siddharthvipul> we are 5 minutes over the time..let's close this
09:04:41 <pingou> +1
09:04:42 <mkonecny> I have one ticket I need help with
09:04:45 <jednorozec> pingou, will see what I can achieve until tomorrow. And talk to you tahn.
09:04:46 <siddharthvipul> we can discuss tomorrow or in other channel
09:04:55 <pingou> jednorozec: sounds good
09:05:00 <mkonecny> #link https://pagure.io/fedora-infrastructure/issue/7294
09:05:00 <pingou> mkonecny: the anitya one?
09:05:01 <siddharthvipul> #endmeeting