SoylentNews is people
SoylentNews
Last night, Linode restarted beryllium, the chat server, to do emergency maintenance on the host. At the time, I was doing work on creating the new staging and development environment and didn't think much of it at the time at 3am. Not long after, I got notification that the machine was restarting. I went to sleep, and put it out of my mind. What I didn't notice, since this is the first unexpected reboot in awhile was that there was no startup script for the IRCd, due to the migration from CentOS 6 to Ubuntu 22.04. I got messages via Discord and SMS around 2PM EST (that is to say 12 hours later) that the IRC was down, so I got back home as quickly as possible and got it restarted.
I will add "getting someone else beside me onto the backend" on the things to do in the very short term.
~ NCommander
(Score: 3, Interesting) by juggs on Friday June 02, @08:08PM
As IRC Admin it was my oversight that caused this extended outage of IRC. On the old Bery server I had everything IRC related set in cron to automagically come back up following a server restart which proved to work perfectly numerous times.
When setting up the new Bery server IRC functions the prime focus was on getting everything back up, which I did. I also made sure that the start / stop / cron scripts were all functional for the Solanum ircd, Atheme services and various bots. And by functional, I mean they ran fine from the command line.
What I neglected to do however was actually add the overall cron script (cron.all) into the cron table!
NCommander and/or kolie are welcome to contact me on my SN email address if necessary but really a quick look in /home/sylnt/bin/ and /home/ircbots/bin/ should make things clear. cron.all goes in the cron table and calls the other cron.XXX individual scripts. The start. and stop. are used for manual control when required.
Fairly sure I documented this all on twiki or wiki - but.... yeah, good luck with that :/