[BUG] TMC consumes too much RAM - Out Of Memory Exception #39
Description
Context
- Install type -> Baremetal
- Mount type -> Fuse
- CPU -> Ryzen 5 2600X
- RAM -> 32G
- Storage -> 1TB NVMe SSD
- Torbox files -> 74 directories, 292 files
Observations
After some time running, less than 24 hours as shown in the logs, the python process consumes too much memory.
I attached the journalctl -xeu torbox.service output below, anonymized logs in which I've removed the useless lines to enhance its readability.
Since I'm using the systemd config file I've shared in #38 the process gets shut down and then restarts gracefully.
(I just need to manually restart my Plex Media Server container afterwards so it can access the fuse mount again.)
Furhermore, when TMC isn't running my services consume about 2G of memory maximum (I don't have much running) but whenever I run TMC this usage jumps to 12G very fast (under 2 hours) then gets stabilized before reserving more and more memory.
I'm not sure if running TMC through docker instead of running it baremetal would mitigate this memory consumption issue, as docker general kills the whole cgroup and restarts the whole container when the system gets oom.