SPW database is growing at around 8-16MB a minute.
Timeline
Post
Remote status
Context
1
@sun isn't this the same problem that plagued SPC? how can it be prevented?
Replies
11
@noyoushutthefuckupdad investigating
@sun @noyoushutthefuckupdad maybe a bug with the Listen activities?
@lain @noyoushutthefuckupdad what makes you think that? I am gonna do some stats of an hour ago and see what's there.
@sun @noyoushutthefuckupdad because i think spc are the only people who use it
@lain @noyoushutthefuckupdad
well on-disk this includes WAL and stuff so I am guessing I'm accidentally grabbing that because when I grab from the objects table by domain for a 4 hour period, the biggest offender is only misskey.io at 3451 kB for the data column of all that data.
well on-disk this includes WAL and stuff so I am guessing I'm accidentally grabbing that because when I grab from the objects table by domain for a 4 hour period, the biggest offender is only misskey.io at 3451 kB for the data column of all that data.
You have to write a psql query to delete non-essential rows from the activities table, but to delay each select query, so the table doesn't lock. Then vacuum full.
@Humpleupagus @noyoushutthefuckupdad I'm gonna query to evaluate what remote server or user is causing all this.
@sun
I had old notifications from Colleen pop up recently either here or FSE. I donβt remember which of her accounts she posted from. Not sure if thatβs a clue but it is what it is. Hope it either helps or doesnβt in useful fashion
@Humpleupagus @noyoushutthefuckupdad
I had old notifications from Colleen pop up recently either here or FSE. I donβt remember which of her accounts she posted from. Not sure if thatβs a clue but it is what it is. Hope it either helps or doesnβt in useful fashion
@Humpleupagus @noyoushutthefuckupdad
@noyoushutthefuckupdad @sun make the database ro, nobody can post, nothing ever happens, problem solved