2022-01-02 11:09:37 <nmeum> I seem to be no longer able to reach my riscv64 alpine container at nld5-dev1.alpinelinux.org:22204 has the port forwarding maybe stopped working for some reason?
2022-01-02 11:09:40 <nmeum> > ssh: connect to host nld5-dev1.alpinelinux.org port 22204: Host is unreachable
2022-01-02 11:13:22 <ikke> perhaps, let me check
2022-01-02 11:15:13 <ikke> Seems like the ip was changed, so we didn't create a static record
2022-01-02 11:16:30 <ikke> oh, the container is just not started
2022-01-02 11:16:47 <ikke> even though autostart is enabled
2022-01-02 11:17:12 <ikke> ok, started, it should work again
2022-01-02 11:23:11 <nmeum> yes, does work again. thanks! :)
2022-01-02 11:26:21 <ikke> for some reason, all the rv64 containers did not start
2022-01-04 09:32:01 <ikke> clandmeter: what are we going to do about those mirrors? 
2022-01-04 13:32:27 <clandmeter> hi ikke 
2022-01-04 13:32:43 <clandmeter> ive been swamped recent days/weeks
2022-01-04 13:33:27 <clandmeter> so im not sure which issue you are referring too? do we have one on gitlab?
2022-01-04 13:34:46 <ikke> Mirror space that zabbix alerts about 
2022-01-04 20:40:57 <clandmeter> ikke: im not sure
2022-01-04 20:41:07 <clandmeter> we have a few options
2022-01-04 20:41:20 <clandmeter> remove some releases
2022-01-04 20:41:25 <clandmeter> or ask for more storage
2022-01-06 18:59:05 <ikke> clandmeter: I've migrated gitlab traefik to v2.5, and it now forwards ssh
2022-01-06 20:57:55 <nmeum> is the riscv64 builder stuck on zabbix? hasn't uploaded in a while…
2022-01-06 20:58:34 <nmeum> btw: would it be possible to make the riscv64 builder buildlog available before the entire software is build? like on the x86_64 builder
2022-01-06 20:59:04 <nmeum> i.e. the zabbix buildlog currently gives 404 here https://build.alpinelinux.org/buildlogs/build-edge-riscv64/community/zabbix/zabbix-5.4.9-r1.log
2022-01-06 20:59:16 <ikke> It was only possible on x86_64 because that was the host that served distfiles
2022-01-06 20:59:32 <ikke> We don't have any facility yet to upload it live
2022-01-06 20:59:55 <ikke> distfiles has moved, so it's not live anymore for x86_64
2022-01-06 21:00:02 <nmeum> oh, unfourtunate
2022-01-06 21:00:52 <nmeum> always found that useful to figure out if there is a problem with the build, e.g. if it is stuck and such
2022-01-06 21:01:40 <ikke> It's hanging on go
2022-01-06 21:01:57 <nmeum> the riscv64 builder?
2022-01-06 21:02:19 <ikke> yes
2022-01-06 21:02:48 <nmeum> is it hanging on the go 1.7.4 upgrade? because that was a while ago :D
2022-01-06 21:03:50 <ikke> no
2022-01-06 21:04:06 <ikke> I pushed a fix
2022-01-06 21:04:19 <ikke> https://tpaste.us/9P5P 
2022-01-06 21:04:58 <nmeum> ah, hangang on the go parts in zabbix
2022-01-06 21:04:58 <nmeum> i see
2022-01-06 21:05:48 <ikke> algitbot: kick master
2022-01-06 21:06:06 <ikke> algitbot: retry master
2022-01-06 21:08:34 <nmeum> nice, ty
2022-01-06 21:08:59 <nmeum> regarding live log files: would it not be possible to setup a http server on the new distfiles server to get live logfiles again?
2022-01-06 21:09:10 <ikke> nmeum: the issue is uploading the logs
2022-01-06 21:09:20 <ikke> this happens now as a separate step after building the package
2022-01-06 21:10:29 <nmeum> ah
2022-01-06 21:10:30 <nmeum> hm
2022-01-06 21:14:19 <ikke> I'm a bit puzzled
2022-01-06 21:15:52 <ikke> So on most arches we have a plugin for buildrepo that uploads the logs and anounces it on mqtt
2022-01-06 21:16:11 <ikke> but not on x86 and x86_64, yet it it's still working
2022-01-06 21:17:53 <nmeum> hehe
2022-01-06 21:50:59 <nmeum> alternative idea: should it not be possible to write log files from the builders to some sort of nfs and have that nfs mounted on the server powering buildlogs.alpinelinux.org for live buildlog updates? thereby avoiding the upload step altogether?
2022-01-07 07:24:18 <nmeum> is the riscv64 builder actually doing stuff on numpy or is it stuck again?
2022-01-07 10:19:11 <vkrishn> this could be used to deter regular spaced text spams - https://tpaste.us/lyLn 
2022-01-07 14:10:55 <nmeum> ikke: if you have a free minute: could you check what the riscv64 builder is doing on numpy? seems weird that it spends so many hours building it or am I just impatient? :D
2022-01-07 14:26:33 <ikke> You can kick it yourself and get the buildlog
2022-01-07 16:33:10 <ikke> algitbot: kick risv64
2022-01-07 16:33:12 <ikke> algitbot: kick naster
2022-01-07 16:33:15 <ikke> algitbot: kick master
2022-01-07 16:33:49 <Hello71> kick ikke's keyboard?
2022-01-07 21:04:45 <ikke> heh, that's fast, just sent them an e-mail :)
2022-01-12 05:41:29 <ikke> security anouncement for gitlab
2022-01-12 07:26:16 <danieli> ikke: some of the fixes seem especially important, i was notified through a thing at work
2022-01-12 07:30:39 <ikke> building new versions
2022-01-12 11:25:48 <ikke> Upgraded gitlab-test to 14.4.5
2022-01-12 19:10:37 <ikke> gitlab upgraded to 14.4.5
2022-01-14 17:27:41 <nmeum> are we using gitlab 11.3? is there any chance we could enable interactive web terminals for the ci? should ease debugging build/check failures, see https://docs.gitlab.com/ee/ci/interactive_web_terminal/
2022-01-14 17:28:05 <ikke> we are at 14.4 atm :)
2022-01-14 17:28:24 <ikke> It requires things to be setup correctly
2022-01-14 17:28:33 <ikke> and
2022-01-14 17:28:40 <ikke> it does not work with docker executors afaik
2022-01-14 17:29:42 <ikke> "The docker executor does not keep running after the build script is finished. At that point, the terminal automatically disconnects and does not wait for the user to finish."
2022-01-14 17:59:46 <nmeum> unfourtunate
2022-01-15 13:42:12 <ikke> mps: you still have access to netbox, right?
2022-01-15 13:57:37 <mps> I think so, only not sure if I have password
2022-01-15 14:10:54 <ikke> You do have an account there, so if I need to reset your pw, let me know
2022-01-15 14:17:22 <mps> I don't often connect there
2022-01-15 14:19:30 <mps> but probably would be good to have password if I need to look something or change/add
2022-01-15 14:19:44 <ikke> Sent via pm
2022-01-15 14:20:38 <mps> thanks
2022-01-20 05:48:54 <ikke> clandmeter: mips64 is back at the moment
2022-01-20 09:28:24 <clandmeter> ok, but it cant upload iirc
2022-01-20 09:30:16 <ikke> Correct, its constantly trying 
2022-01-20 22:40:05 <mps> ikke: I cleaned my lxc containers, they are ready for removal
2022-01-21 13:53:40 <danieli> did the lounge die for some hours?
2022-01-21 15:05:11 <clandmeter> danieli: it did
2022-01-21 15:05:18 <danieli> i see
2022-01-21 16:45:27 <ikke> clandmeter: when you have time, can you check usa5?
2022-01-21 16:45:48 <ikke> /var/lib/docker has 50G data which I cannot account for
2022-01-23 18:28:49 <ikke> :.'
2022-01-23 18:28:51 <ikke> :/
2022-01-24 13:55:11 <clandmeter> ikke: are we getting hammered? 
2022-01-24 20:55:12 <ikke> clandmeter: doesn't look like it
2022-01-24 20:55:34 <ikke> load is reasonable on the host
2022-01-24 20:55:37 <ikke> not a lot of requests
2022-01-24 20:56:11 <ikke> restarted the workers
2022-01-25 12:28:17 <mps> I tried to get my password on https://meta.alpinelinux.org/forgot but it says 'No account found with this email address.' for mps@arvanta.net though I receive mails
2022-01-25 12:29:15 <mps> how can I unsubscribe
2022-01-25 12:32:14 <ikke> Send an email to ~alpine/devel+unsubscribe
2022-01-25 12:32:29 <ikke> @lists.alpinelinux.org
2022-01-25 12:32:36 <mps> ikke: thanks
2022-01-25 12:32:39 <ikke> Same for other lists 
2022-01-25 20:00:03 <ikke> clandmeter: ping
2022-01-25 20:00:10 <clandmeter> ping
2022-01-25 20:00:14 <ikke> pong
2022-01-25 20:00:22 <clandmeter> i mean pong
2022-01-25 20:00:25 <ikke> :)
2022-01-25 20:00:25 <clandmeter> was a typo :)
2022-01-25 20:00:32 <ikke> I assumed so
2022-01-25 20:00:45 <ikke> so on usa5, we have the contents of /srv duplicated
2022-01-25 20:00:59 <ikke> one time on vg0 and one time on vg1
2022-01-25 20:01:30 <ikke> I think docker already uses the directory before vg1/storage was mounted on /srv
2022-01-25 20:01:52 <ikke> So there was 60G of space I could not locate
2022-01-25 20:02:16 <ikke> The issue is, I don't really know how to properly merge it
2022-01-25 20:02:36 <ikke> It's not only data, it's the entire structure that is nearly duplicated
2022-01-25 20:03:00 <ikke> I have mounted vg1/storage on /mnt/storage
2022-01-25 20:03:04 <ikke> (and stopped docker)
2022-01-25 20:03:24 <clandmeter> im looking but dont understand yet :)
2022-01-25 20:03:35 <ikke> compare /srv with /mnt/storage
2022-01-25 20:05:20 <clandmeter> ah now i start to understand
2022-01-25 20:05:41 <clandmeter> currently vg1 was not used?
2022-01-25 20:05:47 <ikke> it was mounted on /srv
2022-01-25 20:05:55 <ikke> so hiding /srv on vg0
2022-01-25 20:06:02 <clandmeter> nod
2022-01-25 20:06:35 <clandmeter> so why not just kill it?
2022-01-25 20:06:43 <ikke> kill what?
2022-01-25 20:06:55 <clandmeter> srv data on vg0?
2022-01-25 20:07:05 <ikke> because I think the builder has been writing to there
2022-01-25 20:07:18 <ikke> it has 60G of data
2022-01-25 20:07:33 <ikke> the rv64 builder
2022-01-25 20:08:41 <clandmeter> but if you mount it on another disk the other data is not accessible right? 
2022-01-25 20:08:54 <ikke> correct, unless a process already has it open
2022-01-25 20:09:05 <ikke> so if docker happens to start before the volume is mounted
2022-01-25 20:11:12 <clandmeter> hmm
2022-01-25 20:16:13 <clandmeter> if you ask me the data in /srv is just old
2022-01-25 20:16:19 <MY-R> the volume isnt in fstab?
2022-01-25 20:16:28 <clandmeter> and the disk probably got full
2022-01-25 20:16:39 <clandmeter> and created a new volume and copied the data
2022-01-25 20:16:43 <clandmeter> then mounted it
2022-01-25 20:16:44 <ikke> MY-R: it is
2022-01-25 20:17:04 <MY-R> so how docker can start before localmount service finish? :\
2022-01-25 20:17:07 <ikke> clandmeter: it was vg0 that was getting full lately
2022-01-25 20:17:13 <clandmeter> nod
2022-01-25 20:17:22 <ikke> so something was writing to vg0
2022-01-25 20:17:38 <clandmeter> yes its root
2022-01-25 20:17:44 <ikke> and the only things that take space is /var/lib/docker and /srv
2022-01-25 20:18:07 <clandmeter>  / is mounted on vg0 right?
2022-01-25 20:18:10 <ikke> yes
2022-01-25 20:18:19 <clandmeter>  /dev/mapper/vg0-lv_root on / type ext4 (rw,relatime)
2022-01-25 20:18:37 <clandmeter> so that volume was getting full 
2022-01-25 20:18:56 <ikke> and builds failed because of disk usage
2022-01-25 20:19:11 <ikke> so the builder was definitely writing to vg0-lv_root
2022-01-25 20:19:18 <clandmeter> in the past yes
2022-01-25 20:19:22 <ikke> no
2022-01-25 20:19:24 <ikke> today
2022-01-25 20:19:36 <clandmeter> yes
2022-01-25 20:19:40 <clandmeter> but not in srv i think
2022-01-25 20:19:48 <clandmeter> probably in docker 
2022-01-25 20:19:50 <ikke> why would it fail with disk usage?
2022-01-25 20:20:02 <clandmeter> the containers and such are on vg0
2022-01-25 20:20:13 <clandmeter> probably somethign is filling up the container itself?
2022-01-25 20:20:54 <clandmeter> -rw-r--r-- 1 dnsmasq dnsmasq 523K Jun  5  2021 /srv/compose/aports-build/storage/packages/main/riscv64/APKINDEX.tar.gz
2022-01-25 20:20:57 <ikke> right
2022-01-25 20:21:04 <clandmeter> thats the date i did the change (i think)
2022-01-25 20:21:06 <ikke> /var/lib/docker/overlay2/d6fdd8b5997463d6c4e324f771965705ffb42c1beeaef5fe2730ab4d2cb83ee0/diff/home/buildoze
2022-01-25 20:21:11 <ikke> /var/lib/docker/overlay2/d6fdd8b5997463d6c4e324f771965705ffb42c1beeaef5fe2730ab4d2cb83ee0/diff/home/buildozer
2022-01-25 20:21:34 <ikke> so if we destroy the container, the packages are gone?
2022-01-25 20:21:58 <ikke> no, packages is empty
2022-01-25 20:22:03 <clandmeter> no i dont think so
2022-01-25 20:22:07 <clandmeter> look at the compose
2022-01-25 20:22:23 <clandmeter> some parts are bind mouinted to ./storage
2022-01-25 20:22:32 <ikke> .cache 14G
2022-01-25 20:22:40 <clandmeter> yeah
2022-01-25 20:22:44 <ikke> go 5G
2022-01-25 20:22:46 <clandmeter> i didnt mount all of it
2022-01-25 20:22:53 <ikke> yeah, ok, that makes sense
2022-01-25 20:23:21 <clandmeter> we could if you want
2022-01-25 20:23:30 <clandmeter> if you want to reuse cache parts
2022-01-25 20:23:47 <ikke> Not sure
2022-01-25 20:24:13 <clandmeter> well in this way its also useless
2022-01-25 20:24:33 <clandmeter> in a perfect world the container would be spin for each package cleaning it self up after
2022-01-25 20:25:22 <clandmeter> i think we also discussed about moving it out of docker
2022-01-25 20:25:37 <ikke> we have an lxc container on nld5 already
2022-01-25 20:25:56 <ikke> oh, that's 3.15 I think
2022-01-25 20:26:12 <ikke> yeah, was a test to bootstrap 3.15
2022-01-25 20:27:23 <clandmeter> we can move it
2022-01-25 20:27:34 <clandmeter> but does the destination have the space needed?
2022-01-25 20:28:49 <ikke> vg1 has enough space
2022-01-25 20:29:02 <clandmeter> on nld5?
2022-01-25 20:29:05 <ikke> oh
2022-01-25 20:29:32 <ikke> /var/lib/lxc has 129G
2022-01-25 20:29:37 <clandmeter> if i understand correctly, for usa5 we can just rm -rf /srv (if not mounted) :)
2022-01-25 20:29:59 <clandmeter> nld5 is equinix?
2022-01-25 20:30:00 <ikke> ok, feel free :P
2022-01-25 20:30:02 <ikke> clandmeter: yes
2022-01-25 20:30:35 <clandmeter> ah that the dev box
2022-01-25 20:30:38 <ikke> yes
2022-01-25 20:31:09 <clandmeter> ok, what do we gain if we move it to nld5?
2022-01-25 20:31:40 <clandmeter> usa5 is a system only used by CI iirc
2022-01-25 20:31:56 <ikke> only thing we would gain is that it uses the same structure as other builders
2022-01-25 20:31:57 <clandmeter> its a troubled system which could fail on high load.
2022-01-25 20:32:17 <clandmeter> at least it did that long time ago
2022-01-25 20:33:27 <clandmeter> if we start doing releases for riscv the space is going to be a bottleneck again
2022-01-25 20:33:31 <ikke> yes
2022-01-25 20:33:45 <clandmeter> it would be nice if we could abuse usa5 for that
2022-01-25 20:33:59 <ikke> we could start using it with lxc again
2022-01-25 20:34:10 <clandmeter> but then again, we said we only do stables if we have the hw
2022-01-25 20:34:17 <ikke> nod
2022-01-25 20:34:29 <ikke> which I think is still quite a while out
2022-01-25 20:35:08 <clandmeter> adding lxc to usa5 sounds like the best plan
2022-01-25 20:35:17 <clandmeter> not sure how much trouble it will make
2022-01-25 20:36:16 <clandmeter> how would you manage the CI instances?
2022-01-25 20:36:27 <clandmeter> via qemu or run docker and lxc together?
2022-01-25 20:37:07 <ikke> afaik docker and lxc can run together without issue
2022-01-25 20:38:00 <clandmeter> do you have it on your bucketlist? :)
2022-01-25 20:38:20 <ikke> Runing lxc and docker on one system?
2022-01-25 20:38:23 <ikke> Yes, ofcourse!
2022-01-25 20:38:36 <clandmeter> :)
2022-01-25 20:39:00 <ikke> I think I've even already ticked that one
2022-01-25 20:39:18 <clandmeter> if you can get lxc and docker running on usa5, i dont mind moving the builder to it
2022-01-25 20:39:28 <ikke> try ls /var/lib/docker on nld3 :)
2022-01-25 20:40:33 <clandmeter> but no daemon running :)
2022-01-25 20:40:38 <ikke> no, we stopped it again
2022-01-25 20:40:42 <ikke> but it did run
2022-01-25 20:40:52 <clandmeter> sure it can run
2022-01-25 20:40:57 <ikke> but?
2022-01-25 20:40:59 <clandmeter> but without issues preferabel :)
2022-01-25 20:41:07 <ikke> what kind of issues?
2022-01-25 20:41:14 <clandmeter> without network madness
2022-01-25 20:41:30 <clandmeter> but i think awall has some logic to handle it now
2022-01-25 20:41:37 <ikke> As long as docker has it's own subnet, it should not be an issue
2022-01-25 20:41:41 <ikke> yes
2022-01-25 20:41:54 <clandmeter> but i have not looked in it for a long time
2022-01-25 20:41:58 <ikke> what we cannot do easily afaik is let docker use part of the dmvpn subnet space
2022-01-25 20:42:06 <ikke> but I don't think that matters
2022-01-25 20:42:09 <clandmeter> i think ncopa also had docker running in lxc iirc
2022-01-25 20:42:23 <ikke> right, but with a lot of extra privileges I think
2022-01-25 20:42:32 <ikke> I tried it, but could not get it quite running
2022-01-25 20:42:37 <clandmeter> yes
2022-01-25 20:42:47 <clandmeter> but running docker on the host is that same issue i guess
2022-01-25 20:42:57 <ikke> what same issue?
2022-01-25 20:43:07 <ikke> afaik it runs ootb
2022-01-25 20:43:15 <clandmeter> you need to give it perms to run it in lxc
2022-01-25 20:43:30 <ikke> right, and provide the cgroup interface
2022-01-25 20:43:32 <clandmeter> but thats the same as running it on the host itself
2022-01-25 20:43:39 <ikke> yeah, basically
2022-01-25 20:43:48 <clandmeter> it would not be for security reasons to do it
2022-01-25 20:44:02 <ikke> indeed, so I don't see the advantage
2022-01-25 20:44:04 <clandmeter> could be practical, but im not sure.
2022-01-25 20:44:18 <ikke> a bit like docker in docker
2022-01-25 20:44:38 <clandmeter> not even sure how iptables would work
2022-01-25 20:45:02 <clandmeter> so if you are up to it, please go ahead and add lxc on usa5
2022-01-25 20:45:21 <ikke> I've enabled awall on deu2 (gitlab host)
2022-01-25 20:45:30 <ikke> as long as you have the dedicated chains enabled
2022-01-25 20:45:31 <clandmeter> then i will try to reserve some time to move it (when my house is in a state of less attention)
2022-01-25 20:46:19 <clandmeter> or you can also try it, if you can find the time. all fine by me.
2022-01-25 20:47:19 <clandmeter> ill go and watch some tv now to relax a bit. i feel i need it :)
2022-01-25 20:47:24 <ikke> o/
2022-01-25 20:47:32 <ikke> so regarding the builder
2022-01-25 20:47:39 <ikke> I probably can restore it for now?
2022-01-25 20:47:40 <clandmeter> let me know what you decide
2022-01-25 20:47:47 <ikke> I mean, in the interrim
2022-01-25 20:47:51 <clandmeter> sure
2022-01-25 20:47:53 <ikke> 9G free space
2022-01-25 20:47:59 <clandmeter> i think you can kill /srv
2022-01-25 20:48:03 <ikke> ok, will do that
2022-01-25 20:48:11 <clandmeter> if that helps 
2022-01-25 20:48:12 <ikke> or I can move it asside even
2022-01-25 20:48:18 <ikke> oh no
2022-01-25 20:48:23 <ikke> that's on a different partition
2022-01-25 20:48:37 <ikke> Have a good evening
2022-01-26 07:08:40 <ikke> clandmeter: we are not using tls-alpn-01 for letsencrypt, right? 
2022-01-26 07:09:16 <ikke> Afaik we only use dns-01
2022-01-26 07:38:27 <clandmeter> uhm, we are mostly using the wildcard dns cert
2022-01-26 07:39:20 <clandmeter> not sure if we have some service that fetches one itself
2022-01-26 07:44:27 <ikke> But that would most likely use an http challenge 
2022-01-26 08:42:22 <clandmeter> yes it will
2022-01-26 08:42:32 <clandmeter> what is it about alpn?
2022-01-26 08:50:10 <ikke> They found issues with the implementation and are revoking certs
2022-01-26 08:50:44 <ikke> https://community.letsencrypt.org/t/2022-01-25-issue-with-tls-alpn-01-validation-method/170450
2022-01-26 20:04:41 <peer> dns-01 is best
2022-01-26 20:06:14 <psykose> wildcard on LE requires dns-01 so this was unaffecting if everything was using a wildcard
2022-01-26 20:06:26 <psykose> (unless they changed it in the past 3 months..)
2022-01-26 20:14:13 <ikke> Yes, and tls-alpn-01 is fairly new
2022-01-26 20:14:26 <ikke> We use wildcard for most, but there might be a dns-01 cert here and there
2022-01-26 20:15:01 <psykose> uhh
2022-01-26 20:15:16 <psykose> do you get your wildcard from somewhere not-LE
2022-01-26 20:16:11 <psykose> (because LE wildcards require dns-01 so those words confused me)
2022-01-26 20:22:14 <rails> wildcards are dns-01 only :D
2022-01-26 20:23:15 <ikke> psykose: we use wildcars, therefore dns-01 in most situations
2022-01-26 20:44:09 <vkrishn> not a great option, log folder can be squash`ed fs and remounted(overlayfs) there itself
2022-01-26 20:44:17 <vkrishn> would save tons of space
2022-01-26 20:44:22 <ikke> vkrishn: wrong channel?
2022-01-26 20:44:41 <vkrishn> nope, just toooo late response
2022-01-26 20:44:43 <ikke> ah
2022-01-26 20:45:03 <ikke> What was that a response to?
2022-01-26 20:45:19 <vkrishn> logs hogging the spaces in servers
2022-01-26 20:51:51 <vkrishn> ikke: https://tpaste.us/eV1X
2022-01-26 20:56:04 <ikke> oh, heh
2022-01-26 21:00:02 <ikke> How would squashfs help for a log directory? It's read-only?
2022-01-26 21:01:26 <vkrishn> overlayfs
2022-01-26 21:01:48 <ikke> yes, but what would squashfs add?
2022-01-26 21:02:45 <ikke> We want these logs continuously rotating, not keeping them indefinitely
2022-01-26 21:03:05 <vkrishn> this would be ideal in situation where there only logs are still needed, else it can be purged or backed'up
2022-01-26 21:03:26 <vkrishn> I where old logs are still needed
2022-01-26 21:03:35 <ikke> That's not the case here
2022-01-26 21:03:36 <vkrishn> I mean
2022-01-26 21:03:41 <vkrishn> ok
2022-01-27 16:55:34 <ikke> clandmeter: I have built gitlab 14.5 and deployed it on gitlab-test
2022-01-27 17:37:32 <ikke> I've cleaned up the .cache dirs on the builders
2022-01-29 15:35:03 <ikke> clandmeter: could you clean up your container? clandmeter-edge-x86_64 73G :-)
2022-01-29 17:16:31 <ikke> If you have time, could you also review https://gitlab.alpinelinux.org/alpine/infra/mirror-status/-/merge_requests/6
2022-01-31 18:42:07 <ikke> clandmeter: another MR, this one is just an esthetic change
2022-01-31 18:42:09 <ikke> https://gitlab.alpinelinux.org/alpine/infra/mirror-status/-/merge_requests/5