| === JanC_ is now known as JanC | ||
| ironhalik | Hello | 10:54 |
|---|---|---|
| ironhalik | got a small issue with xenial server, and dont really know where to start debugging it | 10:55 |
| ironhalik | most of my log files in /var/log are empty and not logged to, except the ones rotated by logrotate | 10:55 |
| ironhalik | ie. haproxy.log is empty, while haproxy.log.1 is actively written to | 10:56 |
| ironhalik | rsyslog is configured to log to haproxy.log, and logrotate uses default config | 10:56 |
| k_sze_ | Does unattended-upgrade refrain from automatically rebooting if it can't send mail notification? | 13:50 |
| ikonia | I don't believe so | 13:50 |
| ikonia | I believe mail notifications is just a "bonus" feature, nothing has a hard depend on it | 13:51 |
| ikonia | I could be wrong, | 13:51 |
| ikonia | I certainly have used it without an MTA being defined and never had a problem | 13:51 |
| k_sze_ | I somehow don't get automatic reboots. :( | 13:52 |
| ikonia | isn't it an option you have to enable ? | 13:52 |
| ikonia | (I don't know if it's on by default) | 13:52 |
| ikonia | I'm working from memory and I think there was an option to enable auto reboot | 13:53 |
| k_sze_ | There is, and I have enabled it in the config. | 13:53 |
| ikonia | is the upgrade not acutally completing perhaps ? | 13:54 |
| ikonia | do the logs show a sucessful upgrade ? | 13:54 |
| k_sze_ | Which log should I look at? If I ssh into the server, I see the message "*** System restart required ***" | 13:54 |
| ikonia | interesting so that suggests it knows it needs the resetart but doesn't execute it | 13:55 |
| k_sze_ | But the /var/run/reboot-required file is quite old. | 13:55 |
| ikonia | there should be an upgrade log | 13:55 |
| ikonia | is the /var/run/reboot-required file actually overwritable ? | 13:55 |
| ikonia | try removing it and re-running the upgrade | 13:55 |
| k_sze_ | It is overwritable | 13:57 |
| ikonia | then why isn't it being overwriteable | 13:57 |
| ikonia | I wonder...does it need a reboot before it can start ? | 13:58 |
| ikonia | has the upgrade actually happened, or is it the prep thats failing | 13:58 |
| ikonia | (if you say the reboot-required file is old) | 13:58 |
| k_sze_ | Well, the reboot-required is dated Sept 3 | 13:58 |
| k_sze_ | this month | 13:58 |
| ikonia | when did you run the upgrade ? | 13:58 |
| k_sze_ | You mean the last time I ran updates manually? | 14:02 |
| k_sze_ | what the.... | 14:14 |
| k_sze_ | if I run `unattended-upgrade -v -d` manually in the terminal as root, it *does* reboot. | 14:14 |
| k_sze_ | But I still don't get an e-mail notification. :( | 14:15 |
| k_sze_ | Even though I know that root's .mailrc works. I *have* tried to send a test message using s-nail. | 14:15 |
| tomreyn | does it run as root, though? | 14:30 |
| k_sze_ | tomreyn: erm... why would it not run as root? | 14:34 |
| k_sze_ | Actually, how does unattended-upgrade get run on Xenial server? | 14:38 |
| k_sze_ | Using a systemd timer? | 14:40 |
| tomreyn | according to /etc/cron.daily/apt-compat that's how | 14:44 |
| tomreyn | systemctl list-timers lists apt-daily.timer and apt-daily-upgrade.timer. i thinkt he former is what triggers u-u | 14:47 |
| hallyn | cpaelzer: rbasak: ok this is funky. With zesty and artful vms, using uvtool... I usually pass in a --run-script-once which does som eupdating and then a reboot (then writes over nc to a socket to tell me script it's ready). | 15:42 |
| hallyn | but with zesty and beyond, the vm hangs on reboot until i do a virsh console. as soon as i do that, it proceeds. wtf? | 15:42 |
| === demon_spork is now known as demonspork | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!