This is needed for restbase as parsoid seems to be using 0.8gb on misc3.
And also electron + mathoid being on there.
But for now we only need 2gb of ram and 2 cpu.
This is needed for restbase as parsoid seems to be using 0.8gb on misc3.
And also electron + mathoid being on there.
But for now we only need 2gb of ram and 2 cpu.
It seems I’ve managed to reduce the ram useage by restbase allowing it to start on misc3 now :)
But nothing is hitting it yet so it shouldent be using ram restbase side (except the node script)
Seems this is needed now, looking at https://ganglia.miraheze.org/?r=hour&cs=&ce=&c=Miraheze&h=misc3.miraheze.org&tab=m&vn=&hide-hf=false&m=load_one&sh=1&z=small&hc=4&host_regex=&max_graphs=0&s=by+name it's very close to using 1gb of ram and also when puppet runs and restarts the services when the config changes, restbase and mathoid fail to start and in syslog i see a bunch of "acct.service: Failed to set invocation ID on control group /system.slice/acct.service, ignoring: Operation not permitted" (same for all the other services too)
Restarting the services manually after puppet ran works, so this means it reached the ram limit as puppet uses ram when it runs.
I think that I have permissions to do the upgrade, but since we're waiting on @labster for funds we will wait to see if @Southparkfan replies before then.
Since ganglia is unavailable I cannot see information regarding server usage now. No approval (yet).
Every day mathoid seems to stop working due to a OOM so i have to restart it. (same for restbase)
Somehow I prefer the ganglia graphs, but that is irrelevant for this particular task; approved.
Keeping assigned to myself for actual upgrade.
I would like to know what I'm approving first, thanks. Is this 1GB SVZS to 2GB SVZS?
Reassigning to SPF, as he asked to be assigned, and yet someone else reassigned like an hour later.