:icon_update: Pleroma 2.0 and some performance tweaks
We updated to Pleroma 2.0 (and i fixed some db schema issues i head 
).
Also instead of using a single node for pleroma and database, we now have a dedicated db server.
Everything on a single node ate too much CPU and RAM resources.
I might transfer over all my services at one point.
).Also instead of using a single node for pleroma and database, we now have a dedicated db server.
Everything on a single node ate too much CPU and RAM resources.
I might transfer over all my services at one point.
:icon_update: Pleroma 2.0 and some performance tweaks
@kura how much cpu and RAM does they eat？ 
:icon_update: Pleroma 2.0 and some performance tweaks
@mashiro
it is a tripple core node, and pleroma postgres connections sometimes ate 100% on all 3 cores for a while.
also the node only has ~4gb ram. and the current db node says it is using 50% [2GB] right now.
it is a tripple core node, and pleroma postgres connections sometimes ate 100% on all 3 cores for a while.
also the node only has ~4gb ram. and the current db node says it is using 50% [2GB] right now.
:icon_update: Pleroma 2.0 and some performance tweaks
@kura OMG，it really eats a lot 
:icon_update: Pleroma 2.0 and some performance tweaks
@mashiro
that is most likely a result of some dubious stuff i did in the past with pleroma, biggest issue is the usage of develop i guess
that is most likely a result of some dubious stuff i did in the past with pleroma, biggest issue is the usage of develop i guess
:icon_update: Pleroma 2.0 and some performance tweaks
@kura Congratulations 
:icon_update: Pleroma 2.0 and some performance tweaks
i was switching to develop, which was stable. and at that point i only had myself and a friend on this server.
now that other people are joining, i am more for the stable codebase instead of bleeding edge. thats why i took the time with sucha long downtime and update to latest stable (was behind quite a lot on develop bc i waited for stable to catch up).
tho there were some unplanned issues during upgrade. and hardware limitations after the upgrade. now everything is running smoothly. and more stable as well thanks to extra db server.