#archlinux32 | Logs for 2019-07-11

Back
[00:38:01] -!- samantaz has quit [Ping timeout: 268 seconds]
[00:39:50] -!- samantaz has joined #archlinux32
[01:11:50] -!- thePiGrepper has joined #archlinux32
[02:16:52] -!- MrBIOS has quit [Quit: MrBIOS]
[02:30:23] -!- sadek has joined #archlinux32
[02:30:45] <sadek> Hello. I just wanted to say thank you for your work. My machine is working like nothing before =)
[02:33:28] <tirsek> \o/
[02:35:24] -!- tirsek has quit [Quit: Time for a reboot]
[02:39:49] -!- tirsek has joined #archlinux32
[02:45:54] -!- sadek has quit [Quit: leaving]
[02:55:52] -!- samantaz has quit [Ping timeout: 245 seconds]
[03:01:35] -!- thePiGrepper has quit [Ping timeout: 258 seconds]
[04:26:44] -!- thePiGrepper has joined #archlinux32
[04:30:24] -!- buildmaster has quit [Remote host closed the connection]
[04:31:03] -!- buildmaster has joined #archlinux32
[04:31:04] <buildmaster> !rq buildmaster
[04:31:05] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[04:49:37] -!- thePiGrepper has quit [Ping timeout: 248 seconds]
[05:05:22] -!- thePiGrepper has joined #archlinux32
[05:17:14] -!- thePiGrepper has quit [Ping timeout: 272 seconds]
[05:17:58] -!- thePiGrepper has joined #archlinux32
[05:38:22] -!- thePiGrepper has quit [Ping timeout: 258 seconds]
[06:25:32] -!- deep42thought has joined #archlinux32
[06:25:32] <buildmaster> Hi deep42thought!
[06:25:32] <buildmaster> !rq deep42thought
[06:25:34] <phrik> buildmaster: <deep42thought> please don't go insane - I think I can't fix you like I can fix the buildmaster
[06:25:57] <deep42thought> sadek: thanks for your kind words!
[06:30:12] -!- abaumann has joined #archlinux32
[06:30:12] <buildmaster> Hi abaumann!
[06:30:12] <buildmaster> !rq abaumann
[06:30:14] <phrik> buildmaster: <abaumann> gogo-gadgeto-strace
[06:30:19] <deep42thought> Good morning, abaumann!
[06:30:21] <abaumann> morning deep42thought
[06:31:16] <abaumann> mmh. my eurobuild6 slaves were producing some 'unknown' build states due to missing PGP sign keys..
[06:31:35] <deep42thought> we should add a regex for that
[06:31:46] <abaumann> good point
[06:31:53] <deep42thought> btw: you should consider putting a sync url for your mirror in place
[06:32:17] <deep42thought> e.g. a php or whatever site which syncs the mirror when retrieved
[06:32:29] <deep42thought> and then the build slaves can easily trigger the resync
[06:32:49] <abaumann> aha.
[06:32:52] <deep42thought> because I think, that the haskell mess always starts with a (slightly) outdated mirror
[06:33:17] <abaumann> I do it every 5 minutes (the rsync), but true.. there is a windows for packages being slightly outdated.
[06:33:23] <abaumann> especially if they are small.
[06:33:59] <abaumann> so, a trigger with curl just before calling build-package should be ok?
[06:34:05] <deep42thought> yes
[06:34:10] <deep42thought> there is already one in place
[06:34:15] <abaumann> ah?
[06:34:35] <deep42thought> # commands to run to update the i686 package mirror used for installing i686 packages,
[06:34:35] <deep42thought> # e.g. the one from /etc/pacman.d/mirrorlist32 (useful on build slaves which do not
[06:34:35] <deep42thought> # have the master mirror in the mirrorlist32)
[06:34:35] <deep42thought> #mirror_update_command=''
[06:34:36] * abaumann misses again some (long implemented) features of the build system..
[06:34:49] <abaumann> ah. right.
[06:35:00] <deep42thought> I even commented it properly :-D
[06:35:06] <abaumann> wow ;-)
[06:35:35] <abaumann> yea, says is all :-)
[06:35:46] <deep42thought> this command may also block
[06:35:56] <deep42thought> so it's a good idea to do the sync synchronously
[06:39:23] -!- titus_livius has joined #archlinux32
[07:13:58] -!- nit-picker has quit [Remote host closed the connection]
[07:14:06] -!- deep42thought has quit [Quit: Leaving.]
[07:55:35] -!- deep42thought has joined #archlinux32
[07:55:35] <buildmaster> Hi deep42thought!
[07:55:35] <buildmaster> !rq deep42thought
[07:55:36] <phrik> buildmaster: <deep42thought> db-update --fuck-up
[07:59:02] <deep42thought> buildmaster, phrik: I'm at your disposal - let's `db-update --fuck-up` the buildmaster!
[07:59:54] -!- buildmaster has quit [Remote host closed the connection]
[08:05:19] -!- buildmaster has joined #archlinux32
[08:05:20] <buildmaster> !rq buildmaster
[08:05:21] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[08:05:28] <deep42thought> buildmaster: what's up?
[08:05:28] <buildmaster> up? I'm up for 3 hours, 41 minutes, load average: 0.27, 0.11, 0.03
[08:10:48] <deep42thought> abaumann: dns is changed, there buildmaster.archlinux32.org now points to the new one (buildmaster-new and buildmaster-old point to what they name)
[08:11:21] <abaumann> oh.
[08:11:25] <abaumann> that was fast. :-)
[08:11:33] <deep42thought> it's not yet operational, though
[08:11:38] <abaumann> I'm struggling with build slaves accumulating git commands
[08:11:40] <abaumann> electron4
[08:11:53] <abaumann> I blacklisted this b*hit
[08:12:01] <deep42thought> what do you mean by "accumulating git commands"?
[08:12:12] <deep42thought> there are some git refreshs before building, do you mean those?
[08:12:20] <abaumann> I have 33 git commands running in parallel. :->
[08:12:26] <deep42thought> ohhh
[08:12:51] <abaumann> :-)
[08:13:13] <deep42thought> who starts those?
[08:13:16] <deep42thought> the PKGBUILD?
[08:13:19] <deep42thought> or our build scripts?
[08:13:29] <abaumann> no, I think it's inside electron4 somewhere..
[08:13:35] <deep42thought> O.o
[08:13:42] <abaumann> ..somebody with a very big machine thought it's a good idea..
[08:13:43] <deep42thought> put a lock on git ;-)
[08:14:02] <abaumann> electron never builds in any version, so blacklisting all of them seems fine to me
[08:14:17] <deep42thought> `alias git='flock /tmp/git; git'` or something
[08:14:21] <abaumann> hehe
[08:14:31] <deep42thought> what will that kill?
[08:14:39] <abaumann> I'm fighting Gateway Timeout 502 when synching the mirror..
[08:14:50] <deep42thought> O.o
[08:14:52] <abaumann> ..only to find out, that my 33 gits are competing against one rsync :-)
[08:14:55] <deep42thought> there is no gateway
[08:15:00] <deep42thought> it's a plain nginx
[08:15:16] <deep42thought> or do you mean git.archlinux32.org?
[08:15:21] <abaumann> ah. I thought the message comes from my webserver.
[08:15:26] <abaumann> I think, it's mine.
[08:15:38] <deep42thought> :-D
[08:15:45] <abaumann> now with 33 gits less, the sync should be much faster. :->
[08:15:52] <deep42thought> lol
[08:16:14] <abaumann> ah, 7 seconds :-)
[08:16:44] <abaumann> git clone --branch=69.0.3497.128 --depth=1 https://chromium.googlesource.com
[08:16:46] <phrik> Title: chromium/src.git - Git at Google (at chromium.googlesource.com)
[08:16:49] -!- Dimtree has quit [Read error: Connection reset by peer]
[08:16:50] <abaumann> 33 times
[08:16:51] <abaumann> cool
[08:16:56] <deep42thought> errr
[08:17:05] <deep42thought> this sounds like a task for a caching proxy
[08:17:07] <abaumann> the best one is, if you kill them, they reappear.
[08:17:15] <abaumann> this I consider a DoSA
[08:17:28] <deep42thought> "hmm, that git is still running - maybe I should start another, identical one"
[08:17:41] <abaumann> !grab deep42thought
[08:17:41] <phrik> abaumann: Bazinga!
[08:17:48] <abaumann> lol
[08:17:54] <deep42thought> the question which remains is: why does it start after 33 processes?
[08:18:10] <abaumann> "Build master is not sane.
[08:18:12] <abaumann> oups.
[08:18:24] <deep42thought> that's the new one
[08:18:25] <abaumann> should I stop the slaves?
[08:18:25] <deep42thought> it's ok
[08:18:28] <deep42thought> no
[08:18:30] <abaumann> ok
[08:18:40] <deep42thought> they will do nothing when the build master is not sane
[08:18:41] <abaumann> ah. you are transfering the database
[08:18:44] <deep42thought> it will bail out early
[08:18:47] <abaumann> how sane :-)
[08:18:54] <deep42thought> (that's what this check is mainly intended for)
[08:19:44] -!- nit-picker has joined #archlinux32
[08:19:45] <deep42thought> hmm, I forgot to put the nitpicker into my autostart
[08:20:17] <deep42thought> this one is supposed to run any asynchronous, thorough checks
[08:20:26] <deep42thought> e.g. dependency checks (which are not yet implemented)
[08:20:34] <deep42thought> just fyi
[08:20:34] <abaumann> ah.
[08:20:42] <abaumann> good to know
[08:21:06] <deep42thought> last electron is from 2019-04-12, so: not that old
[08:21:11] <deep42thought> I would not blacklist it for now ...
[08:21:48] <abaumann> electron and electron2 were already on the list, now we got an electron4, and there is no electron3
[08:21:59] <deep42thought> oh
[08:22:03] <deep42thought> now I understand
[08:22:04] <deep42thought> :-D
[08:24:42] <deep42thought> can you update the letsencrypt certs on the new buildmaster, please?
[08:25:13] <abaumann> ok.
[08:26:42] -!- buildmaster has quit [Read error: Connection reset by peer]
[08:26:48] <deep42thought> oops
[08:28:44] <abaumann> ok. should be ok.
[08:28:50] <deep42thought> \o/
[08:29:12] -!- Dimtree has joined #archlinux32
[08:30:04] <abaumann> "Build slave "eurobuild3" is unnknown to the database.
[08:30:07] <abaumann> erm
[08:30:14] <deep42thought> not yet re-emitted ...
[08:30:15] <deep42thought> sry
[08:30:16] <abaumann> /home/master/builder/bin/slave-build-connect: line 51: mysql_run_query: command not found
[08:30:19] <abaumann> oh. :-)
[08:30:22] <deep42thought> I'm still fiddling with the firewall
[08:30:32] <abaumann> easy. :-)
[08:33:01] <abaumann> ok, doing 6 parallel 'git gc' also is not very helpful - doing some cleanup on eurobuild6
[08:35:16] -!- buildmaster has joined #archlinux32
[08:35:16] <buildmaster> !rq buildmaster
[08:35:18] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[08:42:26] <deep42thought> hmm, dbsnapshot on the new buildmaster gives "Access denied for user 'root'@'localhost' (using password: NO)"
[08:42:31] <deep42thought> (run as root)
[08:44:20] <abaumann> same passwords as on the old buildmaster?
[08:44:33] <deep42thought> no .my.conf - as on the old buildmaster
[08:44:37] <abaumann> ah
[08:44:49] <deep42thought> how did it work there?
[08:45:03] <abaumann> dunno
[08:45:08] <deep42thought> :-D
[08:45:13] <deep42thought> it's your script ^^
[08:45:18] <abaumann> oh.
[08:45:19] <abaumann> :-)
[08:45:27] <abaumann> lemme see
[08:53:47] <abaumann> dbsnapshot
[08:53:48] <abaumann> Enter password:
[08:53:50] <abaumann> grmpf
[08:54:09] <abaumann> I'm writing a .my.cnf
[08:54:17] <deep42thought> looks like the new buildmaster accepts built packages :-)
[08:54:35] <deep42thought> yeah, having a .my.cnf is probably the right thing
[08:54:41] <abaumann> nice :-)
[08:55:01] <abaumann> most likley it worked on the old buildmaster, because there was a grant all to root without password..
[08:55:05] <buildmaster> pentium4/dvd+rw-tools are broken (says nlopc43).
[08:55:11] <deep42thought> I doubt that
[08:55:17] <abaumann> and they also break on the new buildmaster :-)
[08:55:21] <deep42thought> because I copied the complete database
[08:55:28] <deep42thought> yeah, nice, isn't it?
[08:55:30] <abaumann> "permissions need to be 600)."
[08:55:31] <abaumann> mmh.
[08:55:47] <abaumann> I doubt _that_.
[08:55:58] <deep42thought> what?
[08:56:08] <abaumann> stackoverflow is more or less a collection of esotheric modern woodoo instead of a source for technical information :-)
[08:56:16] <deep42thought> !grab abaumann
[08:56:17] <phrik> deep42thought: Tada!
[08:56:57] <abaumann> aha.
[08:57:07] <abaumann> better. wrong section. must by mysqldump not mysqld of course
[08:57:10] <abaumann> "mysqldump: Couldn't execute 'SELECT BINLOG_GTID_POS('', '0')': You are not using binary logging (1381)
[08:57:20] <abaumann> this sounds more interesting..
[08:58:09] <deep42thought> like with usual information, there are three levels of spam: syntactic spam, semantic spam and content spam - stackoverflow has lot of the third kind, you say?
[08:58:22] <abaumann> !grab deep42thought
[08:58:23] <phrik> abaumann: Bazinga!
[08:58:24] <abaumann> nicely put.:-)
[08:59:15] <buildmaster> i686/dvd+rw-tools are broken (says nlopc43).
[08:59:44] <abaumann> yeah. dvd_rw-tools I fixed downstream, now it should be fixed upstream for newer glibc, but now the source tarball run away :-)
[09:00:57] <abaumann> show master status is empty
[09:01:26] <deep42thought> where's that?
[09:01:30] <deep42thought> ah, in mysql?
[09:01:45] <abaumann> yep
[09:02:14] <abaumann> show binary logs;
[09:02:15] <abaumann> ERROR 1381 (HY000): You are not using binary logging
[09:02:28] <deep42thought> oh, some config, probably
[09:02:29] <abaumann> did we have log-bin set on the old buildmaster?
[09:02:37] <deep42thought> I guess so
[09:02:39] <deep42thought> let me `diff`
[09:03:06] <abaumann> server.cnf:log-bin=mysql-bin
[09:03:41] <abaumann> The mysql config on the new buildmaster looks pretty vanilla
[09:03:57] * abaumann wonders if "pretty vanilla" qualifies as colour
[09:04:01] <deep42thought> I'll move the old one over
[09:04:05] <deep42thought> !grab abaumann
[09:04:05] <abaumann> ok
[09:04:06] <phrik> deep42thought: Bazinga!
[09:04:12] <deep42thought> !q abaumann
[09:04:12] <phrik> deep42thought: * abaumann wonders if "pretty vanilla" qualifies as colour
[09:04:17] <deep42thought> *phew
[09:04:26] <abaumann> close grabbing. ;-)
[09:07:02] <deep42thought> what's /dev/md-3 supposed to be?
[09:07:12] <deep42thought> swap?
[09:07:26] <deep42thought> dbsnapshot is running!
[09:07:35] <abaumann> cool :-)
[09:07:54] <abaumann> yeah. swap on a software raid
[09:08:00] <abaumann> debateable
[09:08:07] <deep42thought> is the raid really a good idea?
[09:08:14] <deep42thought> why not add two swap partitions?
[09:08:19] <abaumann> well, your swapped data is now really safe :-)
[09:08:30] <abaumann> it makes sense for a laptop for suspend to disk maybe
[09:08:32] <abaumann> otherwise.
[09:08:38] <abaumann> yeah, two swap aeras
[09:08:43] <deep42thought> it's raid1????
[09:08:50] <abaumann> sure:-)
[09:09:01] <abaumann> it's the default Hetzner install script..
[09:09:06] <abaumann> should I change it?
[09:09:11] <deep42thought> yeah, please
[09:09:25] <deep42thought> btw: I noticed, because there are systemd jobs waiting for it
[09:09:41] <abaumann> /dev/md/0 none swap sw 0 0
[09:09:43] <abaumann> /dev/md/3 none swap sw 0 0
[09:09:45] <abaumann> urgh
[09:09:51] <deep42thought> O.o
[09:10:13] <deep42thought> "oh, we have 1G space at the front of the disk, hmm - let's create a swap partition there"
[09:10:44] <abaumann> I should have gone with a manual installation..
[09:10:57] <deep42thought> :-D
[09:11:14] <abaumann> /dev/md/1 /boot ext2 defaults 0 0
[09:11:15] <abaumann> /dev/md/2 / ext4 defaults 0 0
[09:11:31] <abaumann> that's also a little bit dangerous.. considering I remove /dev/md/0 swap now :-)
[09:11:32] <deep42thought> either install automatically and fix all the stuff manually that you don't like (which will be a lot iff you are pedantic, like me) or install manually from the start
[09:11:46] <deep42thought> this is in /etc/fstab?
[09:11:51] <abaumann> yep.
[09:11:56] <deep42thought> put uuids there
[09:12:02] <deep42thought> ah, you don't like uuids
[09:12:02] <abaumann> but I'm fiddling currently there.
[09:12:04] <abaumann> yes. exactly.
[09:13:20] <abaumann> /dev/sda1 4096 67112959 67108864 32G Linux RAID
[09:13:28] <abaumann> first Gigabyte? :-)
[09:14:00] * abaumann coughs nervously
[09:14:32] <deep42thought> I don't understand that layout
[09:14:37] <abaumann> me neither
[09:15:23] <deep42thought> the partitions are "right aligned"?
[09:15:40] <abaumann> hope so.
[09:15:45] <abaumann> I'll have to reboot now..
[09:15:52] <deep42thought> ok
[09:15:54] <deep42thought> go ahead
[09:16:00] <deep42thought> actually, wait a sec
[09:16:24] <abaumann> np
[09:16:33] <deep42thought> ok, fine
[09:16:37] <abaumann> ok
[09:16:50] <abaumann> mmh.
[09:16:54] <abaumann> mdadm.conf
[09:17:03] <abaumann> I think, I fix that one to and regerate the initramdisk
[09:17:23] <deep42thought> why the initrd?
[09:17:43] <abaumann> because I think initrd uses some information from /etc/mdadm.conf when assembling the RAIDs
[09:17:49] <abaumann> or better, I'm not sure anymore..
[09:17:59] <deep42thought> ok, better safe than sorry
[09:19:21] <abaumann> mmh. I stop a raid with mdadm --stop md0 and then it fails one part of the RAID1 and restarts automatically?
[09:19:28] <abaumann> mdadm used to be more stable in the past..
[09:19:52] * abaumann searches for that VNC console on Hetzner..
[09:20:20] <deep42thought> probably not mdadm's fault
[09:20:31] <abaumann> rather admin's fault ;-)
[09:20:33] <deep42thought> on crux, it usually does precisely what I want
[09:20:49] <deep42thought> nah, probably some systemd-auto-fuckup
[09:21:20] -!- buildmaster has quit [Remote host closed the connection]
[09:22:26] <abaumann> yeah. that's the only drawback currently, no KVM
[09:22:35] <abaumann> only the rescue PXE thingy or a VNC/textual installation
[09:22:49] <abaumann> but can be tricked to start a rescued system
[09:23:16] * abaumann rejoices
[09:23:27] <abaumann> md127 : active (auto-read-only) raid1 sda1[0] sdb1[1] 33521664 blocks super 1.2 [2/2] [UU]
[09:23:30] <abaumann> cool
[09:23:32] <deep42thought> yeah
[09:23:50] <abaumann> I might have to nuke the mdraid signature on sda1 and sdb1
[09:24:04] <deep42thought> I had my fileserver running md127 and md126 for years, before I found _where_ to set up that stupid name :-D
[09:24:05] <abaumann> though the GPT type is now "swap" something tries to build a RAID out of it
[09:24:16] <deep42thought> yes
[09:24:29] <deep42thought> done
[09:24:29] <abaumann> mkswap /dev/sda1
[09:24:29] <abaumann> mkswap: cannot open /dev/sda1: Device or resource busy
[09:24:31] <abaumann> grmph
[09:24:44] <abaumann> now
[09:25:00] <deep42thought> I just did a `mdadm --zero-superblock /dev/sda1` - maybe that interfered
[09:25:05] <abaumann> ah. thanks
[09:26:08] * deep42thought feels lucky and starts/enables all systemd services simultanously
[09:30:39] <elibrokeit> So have you played with samurai yet?
[09:30:59] <elibrokeit> Did you setup makepkg.conf to export SAMUFLAGS?
[09:31:14] <abaumann> not yet
[09:31:37] <elibrokeit> Hopefully Pacman 5.2 will have built-in support for SAMUFLAGS...
[09:31:50] <abaumann> oh! that would be cool :-)
[09:32:15] <elibrokeit> I sent in patches. It's no different from. MAKEFLAGS imo
[09:32:33] <abaumann> what's the next one? CARGOFLAGS?
[09:32:45] <elibrokeit> Anyway, I did a mass replacement in my packages to use samurai
[09:32:57] <elibrokeit> Don't joke about cargoflags
[09:33:05] <abaumann> oups.. :-)
[09:33:16] <elibrokeit> My previous patch, before the samu one, was for RUSTFLAGS ;)
[09:33:22] <abaumann> lol
[09:33:57] <elibrokeit> https://patchwork.archlinux.org
[09:33:58] <phrik> Title: [pacman-dev] makepkg: add rust support for *FLAGS and debug-prefix-map - Patchwork (at patchwork.archlinux.org)
[09:37:33] -!- Cthulu201 has quit [Remote host closed the connection]
[09:38:06] -!- Cthulu201 has joined #archlinux32
[09:40:35] <abaumann> 5.102.146.92 buildmaster.archlinux32.org
[09:40:37] <abaumann> in /etc/hosts
[09:40:38] <abaumann> cool
[09:40:46] <abaumann> on eurobuild3
[09:40:49] <abaumann> what was that for?!
[09:40:55] <deep42thought> dunno
[09:41:00] <deep42thought> maybe some dns issues?
[09:41:04] <deep42thought> or from the last move?
[09:41:07] <abaumann> most likely
[09:41:28] <abaumann> so, my mirror sync update script is in place..
[09:44:24] -!- buildmaster has joined #archlinux32
[09:44:24] <buildmaster> !rq buildmaster
[09:44:25] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[09:46:57] <deep42thought> ERROR 1227 (42000) at line 1: Access denied; you need (at least one of) the SUPER, REPLICATION CLIENT privilege(s) for this operation
[09:46:59] <deep42thought> hmmm
[09:47:02] <deep42thought> this is on the slaves
[09:47:10] <deep42thought> (database slaves)
[09:53:32] <abaumann> I'll have to leave soon.. will be back around 4PM o'clockish..
[09:53:38] <deep42thought> ok
[09:53:53] <deep42thought> then I'll get some "real" work done, too in the mean time ;-)
[09:54:05] <abaumann> hehe
[09:54:45] <abaumann> work involving real electrons..
[09:55:03] <abaumann> ok. cu later..
[09:55:04] -!- abaumann has quit [Quit: leaving]
[10:09:29] <deep42thought> $ ping 8.8.8.8
[10:09:29] <deep42thought> ping: socket: Operation not permitted
[10:09:30] <deep42thought> interesting
[10:36:28] -!- deep42thought has quit [Quit: Leaving.]
[10:40:01] * buildmaster resumes sanity.
[10:45:01] <buildmaster> pentium4/kodi is broken (says nlopc43): https://archlinux32.org
[13:09:49] -!- deep42thought has joined #archlinux32
[13:09:49] <buildmaster> Hi deep42thought!
[13:09:49] <buildmaster> !rq deep42thought
[13:09:51] <phrik> buildmaster: <deep42thought> arch32 is on the bleeding edge of arch, which is on the bleeding edge of software?
[13:22:35] <deep42thought> abaumann: you need to update the ssh fingerprint in .ssh/known_hosts on your build slaves
[13:23:58] -!- ofara has quit [Quit: ofara]
[13:26:16] -!- ofara has joined #archlinux32
[14:03:51] -!- ofara_ has quit [Quit: ofara_]
[14:05:33] -!- ofara_ has joined #archlinux32
[14:59:05] <deep42thought> elibrokeit: if you want to reduce the number of connection in offload-build even further, you can do the back-transfer of the built files via `tar -c` inside the ssh and `tar -x` outside the ssh
[14:59:20] <deep42thought> that's what I do
[15:27:20] <buildmaster> girls, my database is dirty again ...
[15:27:20] * buildmaster goes insane.
[16:05:33] * buildmaster resumes sanity.
[16:13:31] -!- deep42thought has quit [Quit: Leaving.]
[16:14:28] <elibrokeit> deep42thought: the design of offload-build was slightly organic, and initially came from an interactive one-liner that just told you where the build file(s) were
[16:15:47] <elibrokeit> I'm not super prioritizing a single connection, but it seemed to make sense to not use a dozen commands in order to copy files to and from -- and I definitively wanted to use srcpkgs anyway, in order to guarantee all available files were copied over.
[16:17:07] <elibrokeit> The problem with emitting tar-c/tar-x on stdout of the ssh command is I'm already playing with stdin/stdout descriptors more than I really like
[17:22:09] -!- abaumann has joined #archlinux32
[17:22:09] <buildmaster> Hi abaumann!
[17:22:10] <buildmaster> !rq abaumann
[17:22:11] <phrik> buildmaster: <abaumann> Never understood why there is a fallback ramdisk, but no fallback kernel image.
[17:24:51] <abaumann> deep42thought: I'm cleaning up the old buildmaster, I just copied the database backups to /data/backup on the new one.
[17:26:31] <abaumann> mmh.. my build slaves are on strike? Maybe I should pay them? ;-)
[17:27:02] -!- deep42thought has joined #archlinux32
[17:27:03] <buildmaster> Hi deep42thought!
[17:27:03] <buildmaster> !rq deep42thought
[17:27:04] <phrik> buildmaster: <deep42thought> ahm, what is actually in the package "vulkan-headers" on i486? I thought, we disabled all vulcans?
[17:27:06] <deep42thought> !grab abaumann
[17:27:07] <phrik> deep42thought: Tada!
[17:27:09] <deep42thought> Hi abaumann!
[17:27:49] <abaumann> hi deep42thought
[17:28:08] <abaumann> things are looking good.. :-)
[17:28:18] <deep42thought> yes
[17:28:31] <abaumann> I will keep the old buildmaster for a while, just in case we missed something
[17:29:13] <deep42thought> yes
[17:29:20] <deep42thought> I already copied a lot of configs ;-9
[17:29:21] <deep42thought> ;-)
[17:30:16] <abaumann> best was again an /etc/hosts entry for buildmaster, now pointing on the old buildmaster to the old buildmaster for the name of the new buildmaster
[17:30:36] * abaumann is somewhat confused about his last sentence himself
[17:35:58] * abaumann moves build logs from old to new buildserver
[17:41:51] <deep42thought> they should be moved already
[17:44:43] <abaumann> where to?
[17:45:00] <abaumann> ah.
[17:45:06] <abaumann> /srv/http/build-logs
[17:45:16] <abaumann> I remember, I made a symlink to /data on the old one.
[17:45:18] <abaumann> So, yes.
[17:45:22] <abaumann> they are moved :-)
[17:46:20] <abaumann> yeah. all this symlinking and mount-binding is bad, just because the old buildmaster had little SSD and bigger platter space.
[17:46:24] <abaumann> this results in a mess.
[17:46:26] <deep42thought> yeah, I needed them to test the new haskell-rescheduling-logic
[17:46:32] <abaumann> no problem
[17:46:45] <abaumann> we have plenty of disk space on the new buildmaster :-)
[17:59:12] <abaumann> must go cooking now.. I think, I'll open a can of sausages (inside joke) :-)
[17:59:31] <deep42thought> lass es dir schmecken!
[17:59:37] <abaumann> danke danke. :-)
[17:59:39] <abaumann> cu
[17:59:41] -!- abaumann has quit [Quit: leaving]
[18:27:35] <buildmaster> pentium4/remind is broken (says eurobuild3): https://archlinux32.org
[18:34:07] -!- isacdaavid has joined #archlinux32
[18:37:43] <buildmaster> pentium4/varnish is broken (says buildknecht): https://archlinux32.org
[18:39:33] <buildmaster> pentium4/gnu-apl is broken (says nlopc46): https://archlinux32.org
[18:41:33] -!- buildmaster has quit [Remote host closed the connection]
[18:41:46] -!- buildmaster has joined #archlinux32
[18:41:46] <buildmaster> !rq buildmaster
[18:41:47] <phrik> buildmaster: <buildmaster> I might be insane, but never confused ... ;-)
[18:54:03] <buildmaster> pentium4/kile is broken (says nlopc46): https://archlinux32.org
[19:10:05] <buildmaster> pentium4/python-warlock is broken (says buildknecht): https://archlinux32.org
[19:21:05] <buildmaster> pentium4/proj is broken (says nlopc43): https://archlinux32.org
[19:21:41] <buildmaster> pentium4/luasec is broken (says nlopc46): https://archlinux32.org
[19:28:22] <buildmaster> pentium4/libbloom is broken (says nlopc46): https://archlinux32.org
[19:32:31] <buildmaster> pentium4/python is broken (says buildknecht2): https://archlinux32.org
[19:42:53] -!- samantaz has joined #archlinux32
[19:56:21] <buildmaster> pentium4/grub is broken (says buildknecht2): https://archlinux32.org
[20:08:24] <buildmaster> pentium4/libraw is broken (says eurobuild6-6): https://archlinux32.org
[20:10:02] <buildmaster> pentium4/guile is broken (says eurobuild6-3): https://archlinux32.org
[20:32:25] <buildmaster> pentium4/xlockmore is broken (says buildknecht2): https://archlinux32.org
[20:36:04] <buildmaster> pentium4/bochs are broken (says eurobuild6-2): https://archlinux32.org
[20:47:47] -!- abaumann has joined #archlinux32
[20:47:47] <buildmaster> Hi abaumann!
[20:47:47] <buildmaster> !rq abaumann
[20:47:49] <phrik> buildmaster: <abaumann> * abaumann wonders if raind is a good name for a system daemon..
[20:48:41] <abaumann> deep43thought: JFYI, bochs is using and old version of packages32 to build, is thus failing (bochs 7144b229de4f8721e919a55511eafb01780ccf05 d57a456faa674c24e8869a26a14c497c95accf1f community), ./get-package-updates -w
[20:48:52] <abaumann> has no effect, says: "Nothing changed."
[20:49:09] <deep42thought> hmmm
[20:49:14] <deep42thought> use seed-build-list, then
[20:49:19] <abaumann> ah, your'there :-)
[20:49:20] <deep42thought> but it should not happen :-(
[20:49:32] <abaumann> the sausages got eaten and were very good :-)
[20:49:43] <abaumann> one can, that is..
[20:49:46] <deep42thought> how were the sausages? (did you eat all 4 of them?)
[20:49:55] <abaumann> not alone. 3 persons. :-)
[20:50:00] <deep42thought> ah, ok :-D
[20:50:06] <abaumann> nice sausages :-)
[20:51:30] <deep42thought> I think some stuff got lost during the transition
[20:51:46] <deep42thought> so there may be some manual fixes necessary now
[20:51:54] <deep42thought> (like for the insanity this afternoon)
[20:52:29] <abaumann> ah, that's an explanation
[20:52:58] <abaumann> -rw-r--r-- 1 root root 38265292 Jul 11 17:54 database-2019-07-11-17:49:20.xz
[20:52:59] <deep42thought> my guess: I missed copying some credentials (e.g. to access the master mirror) which left the buildmaster in an inconsistent state
[20:53:02] <abaumann> -rw-r--r-- 1 root root 27167812 Jul 11 19:56 database-2019-03-14-13:50:36.xz
[20:53:18] <abaumann> the mysqldump also got much smaller.
[20:53:28] <deep42thought> you reversed the date
[20:53:49] <deep42thought> e.g. you ordered by _file_ date
[20:53:49] <abaumann> oh.
[20:53:51] <abaumann> sorry.
[20:53:53] <deep42thought> not by date-of-dump :-)
[20:53:55] <deep42thought> np
[20:54:04] <abaumann> no. that's a very old backup when we did the first migration.
[20:54:08] <abaumann> all ok.
[20:54:20] * abaumann feels sheepish :-)
[20:57:00] <buildmaster> pentium4/jadx is broken (says eurobuild6-5): https://archlinux32.org
[20:58:03] <abaumann> "Error 404: Not Found
[20:58:03] <abaumann> No logs found for this package."
[20:58:05] <abaumann> aha
[20:58:10] <abaumann> lemme check permissions here
[20:58:45] <deep42thought> this is an error generated by the frontend on archlinux32.org
[20:58:50] <deep42thought> not by the buildmaster
[20:58:55] <abaumann> ah.
[20:59:00] <abaumann> right
[20:59:04] <deep42thought> it means, that the log is not (yet?) in the database
[20:59:31] <deep42thought> I should clarify this error message
[21:00:05] <abaumann> naeh. it's clear to me now :-)
[21:00:23] <abaumann> the load on the new buildmaster is much better than on the old one.
[21:00:59] <buildmaster> pentium4/toxcore is broken (says eurobuild6-2): https://archlinux32.org
[21:03:43] <deep42thought> yes, who needs an ssd if you cannot use all cpu cores anyways ;-)
[21:03:59] <deep42thought> plus: now the database can fit more stuff into ram
[21:04:40] <abaumann> mmh. the buildmaster munin scripts are gone. ok, that's something for me to do..
[21:04:51] <deep42thought> yes, I couldn't find them
[21:05:05] <abaumann> lost in action :-)
[21:12:02] <buildmaster> pentium4/budgie-desktop is broken (says nlopc43): https://archlinux32.org
[21:12:07] <deep42thought> moving the buildmaster should not become too painless - otherwise we get crazy ideas (like my son): "Hey, why not move every day?" - "because that will be tedious" - "well, we can just skip one move and directly move to the second-to-next flat!"
[21:15:24] <buildmaster> pentium4/sagemath-doc is broken (says buildknecht): https://archlinux32.org
[21:15:32] <buildmaster> pentium4/consul is broken (says buildknecht2): https://archlinux32.org
[21:17:38] <abaumann> lol
[21:22:02] <buildmaster> pentium4/open-vm-tools are broken (says eurobuild6-1): https://archlinux32.org
[21:29:48] <buildmaster> pentium4/spirv-llvm-translator is broken (says nlopc46): https://archlinux32.org
[21:42:03] <buildmaster> pentium4/lrs are broken (says nlopc43): https://archlinux32.org
[21:57:29] -!- Dimtree has quit []
[21:57:48] <buildmaster> pentium4/pigeonhole is broken (says eurobuild6-4): https://archlinux32.org
[21:59:30] <buildmaster> pentium4/haskell-bitarray is broken (says eurobuild3): https://archlinux32.org
[22:05:42] <buildmaster> pentium4/prometheus are broken (says eurobuild6-2): https://archlinux32.org
[22:21:50] -!- abaumann has quit [Quit: leaving]
[22:30:50] -!- deep42thought has quit [Quit: Leaving.]
[23:10:03] -!- thePiGrepper has joined #archlinux32
[23:19:07] <buildmaster> pentium4/nginx-mainline is broken (says eurobuild6-3): https://archlinux32.org
[23:20:51] <buildmaster> pentium4/lynis are broken (says eurobuild6-2): https://archlinux32.org
[23:28:04] elibrokeit is now known as guys
[23:45:19] <buildmaster> pentium4/xterm is broken (says eurobuild6-2): https://archlinux32.org
[23:53:45] <buildmaster> pentium4/three.js are broken (says eurobuild6-4): https://archlinux32.org