#archlinux32 | Logs for 2022-04-25
Back
[02:26:51] -!- ejjdhfjsu has joined #archlinux32
[06:20:08] -!- volki57 has quit [Remote host closed the connection]
[06:22:20] -!- volki57 has joined #archlinux32
[10:29:31] -!- T`aZ has quit [Remote host closed the connection]
[10:49:37] -!- T`aZ has joined #archlinux32
[10:54:28] -!- T`aZ has quit [Remote host closed the connection]
[11:05:25] -!- T`aZ has joined #archlinux32
[13:16:05] -!- ejjdhfjsu has quit [Remote host closed the connection]
[16:47:18] -!- sunshavi has joined #archlinux32
[19:46:39] -!- titus_livius has joined #archlinux32
[20:21:37] -!- abaumann has joined #archlinux32
[20:21:37] <buildmaster> Hi abaumann!
[20:21:38] <buildmaster> !rq abaumann
[20:21:38] <phrik> buildmaster: <abaumann> in 2019 it seems hard to get simple things working.. like filesystem actually being mounted after a reboot :->
[20:22:15] <abaumann> archlinux32.org is down, I'm trying to resynch the database from the buildmaster. Dropping the buildmaster database seems to last hours and use 2 CPUs non-stop. What the heck is that?
[20:22:52] <abaumann> I thing a 'rm -rf' is at hand now :->
[20:25:15] <abaumann> mysqld_safe cannot be stopped?!
[20:26:46] -!- titus_livius has joined #archlinux32
[20:27:32] <abaumann> rm: cannot remove '/var/lib/mysqltmp/': Device or resource busy
[20:27:41] <abaumann> no mysql is running
[20:27:53] <abaumann> tmpfs /var/lib/mysqltmp tmpfs rw,gid=mysql,uid=mysql,size=100M,mode=0750,noatime 0 0
[20:27:56] <abaumann> nice
[20:28:24] <abaumann> hack around hacks just because the database is not performing ;-)
[20:30:55] <abaumann> oh, kill $(cat /var/lib/mysql/$HOSTNAME.pid)
[20:30:59] <abaumann> is by design..
[20:31:50] <abaumann> having no swap kills mysqld
[20:32:37] <abaumann> So, now I see I/O and only little CPU when importing the database.. this looks more reasonable. :-)
[20:33:39] <abaumann> nice: no changin /etc/fstab requires systemctl daemon-reload.. this is called "improvement"
[20:33:46] <abaumann> *now changing
[20:34:12] <abaumann> tmpdir = /var/lib/mysqltmp
[20:34:23] <abaumann> was missing in /etc/my.cnf.d/server.cnf in mysqld section..
[20:57:47] <abaumann> tmpdir = /var/tmp/
[20:57:50] <abaumann> /dev/md2 2849030168 2252825256 451412008 84% /
[20:57:58] <abaumann> on the buildmater. with 64GB of RAM. :-)
[20:58:03] <abaumann> I'll try to fix that..
[21:01:05] <abaumann> 928 mysql 20 0 4933648 3.3g 5824 S 39.0 86.8 7:31.06 mariadbd
[21:01:13] <abaumann> mmh. now it gets critical on archlinux32.org :-)
[21:01:52] <abaumann> error: cannot lock ref 'refs/heads/master': is at 773631cd686dc5e7f9f4cf25af24b46e4d52cc7c but expected 1cbcfbb29c7759312f50728caea337a7ebe4cfd2
[21:01:58] <abaumann> mmh. this sounds bad.
[21:02:01] <abaumann> on the buildmaster
[21:04:05] <abaumann> ./get-package-updates -w
[21:04:05] <abaumann> Nothing changed.
[21:04:07] <abaumann> mmh. weird.
[21:12:44] <abaumann> so, let's wait.. I leave it on archlinux32.org in a screen
[21:12:46] -!- abaumann has quit [Quit: leaving]
[21:31:07] -!- abaumann has joined #archlinux32
[21:31:07] <buildmaster> Hi abaumann!
[21:31:07] <buildmaster> !rq abaumann
[21:31:08] <phrik> buildmaster: <abaumann> naa.. not testing never has bad consequences.. ;-)
[21:31:24] <abaumann> ok, the import finished on archlinux32.org, but now every query ends in 2xhigh CPU load and timeouts
[21:31:34] <abaumann> and the replication slave is down (is it over tinc or not)
[21:36:23] <abaumann> so, manual repilcation binlog and position repositioning worked, slave is up again.
[21:37:13] <abaumann> let's hope it remains more stable for some time now..
[21:37:14] -!- abaumann has quit [Quit: leaving]
[22:14:33] <girls> abaumann: FYI: there's a script to "re-bootstrap" the db slaves
[22:14:38] <girls> and yes: it's over tinc ;)