load average: 22.17, 32.76, 35.74

Ugh. Upgrading All The Things All At Once On A Resource Constrained Device Might Have Been A Bad Idea.

Managed to SSH in, stop everything, then start them up one by one, and now everything's happy.

@algernon

Yeah, that's happened to me a lot. My latest server, though:

amin@curie:~$ uptime
 11:10:33 up 284 days, 18 min,  1 user,  load average: 0.27, 0.42, 0.48

It just stays like this. The load rarely ticks up at all.

@amin nod

Mine (2014-era mac mini, all original hardware with spinning disk and all) rarely goes above 1, and it only does so because it receives all the logs from iocaine. Most other services are idle most of the time.

But when I upgrade, and everything (postgres, vaultwarden, victorialogs, prometheus, etebase-server, atuin, readeck, ntfy, etc) restarts at the same time, then it becomes a bit unhappy, because the majority of services are suddenly not idle, and there aren't enough cores, memory, nor disk speed to do all this in parallel.

If they were restarted sequentally, the load would never go above 2, and it would take a lot less time. But I'm too lazy to figure out how to get there. O:)

@algernon

I'm hosting five GoToSocial instances and a search engine, though. Among other things. 😆

@amin Speaks a lot about GtS & search engine quality that load remains so low then!

@algernon "I Paid for the Whole CPU!"