You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi,
I ran into memory problems when using this container, and managed constrained it significantly. UISP appears to calculate the number of workers it can use based on how much memory there is, but this is not so accurate on a server that also does other things (like my synology that runs a lot of other services). I discovered you can use the CLUSTER_SIZE variable to set the amount of workers to e.g. 1. It doesn't seem to use the configured memory limit of the container (which I now set to 256MB) but the memory of the host.
Furthermore, the postgresql memory use can be limited a bit, by changing e.g. the shared_buffers=8M, max_parallel_workers=2, changing max_connections didn't appear to work (tried before changing CLUSTER_SIZE variable, and it would not start anymore).
If anyone has more suggestions to limit memory usage, I'm all ears;-)
Kind regards,
Dennis
The text was updated successfully, but these errors were encountered:
This is an excellent observation. I was able to reduce memory usage from 1.6 to 1.3GB using CLUSTER_SIZE=1. Haven't tried postgresql settings yet.
What bothers me however is that there is a process (/usr/local/lib/erlang/erts-10.7.2.3/bin/beam.smp) that takes more than 1GB of RAM on my system. This seems to be a lot for such a tiny application. Can anybody comment what this is used for and if it can be limited?
Hi,
I ran into memory problems when using this container, and managed constrained it significantly. UISP appears to calculate the number of workers it can use based on how much memory there is, but this is not so accurate on a server that also does other things (like my synology that runs a lot of other services). I discovered you can use the CLUSTER_SIZE variable to set the amount of workers to e.g. 1. It doesn't seem to use the configured memory limit of the container (which I now set to 256MB) but the memory of the host.
Furthermore, the postgresql memory use can be limited a bit, by changing e.g. the shared_buffers=8M, max_parallel_workers=2, changing max_connections didn't appear to work (tried before changing CLUSTER_SIZE variable, and it would not start anymore).
If anyone has more suggestions to limit memory usage, I'm all ears;-)
Kind regards,
Dennis
The text was updated successfully, but these errors were encountered: