Docker memory utilization slowly growing due to container log size

[Magic815]

Hi all - 

 

So I'm running UnRAID v6.11.5, and am noticing an odd issue with one of my docker containers. I noticed over the past half a year that my docker memory utilization had been slowly but steadily creeping up from ~50% to 87% with no new docker containers being added to my server. When I hit "Container Size" in the GUI on the Docker tab at the point it reached 87%, I realized that Plex-Meta-Manager was showing ~16GB used in the 'Log' column. After some quick Google searching (and searching this forum) it sounded like it could be something misconfigured with that container, or some corruption that was causing the docker.img to get slowly flooded with logs from that container.

 

So a few days ago, I ended up completely removing that container/image and started with a fresh one. With the new container up and running, the log size initially went back to 0 KB. However, the way I use this container is at 6:00am every day, it runs to update the collections and overlays of my Plex libraries. I'm already recognizing a pattern where each day (i.e. each 6:00am run) the log size grows by ~28.3MB. I'm on day 3, and so the log size is currently back to 82.8MB (see screenshot below). But I suspect it is just going to keep growing steadily until it eventually caps my docker utilization if I don't find a proper fix.

 

I've attached my diagnostics file. Is there anything that looks off in my configuration of that docker container that may explain why the space consumed by logs just slowly gets more and more consumed after each daily run? (I've checked with other UnRAID users in the Plex-Meta-Manager discord server, and none of them report seeing slowly inflating log sizes).

 

(As an aside, InfluxDB seems high too at 2.80GB, but I don't think I've noticed a pattern yet on that one growing steadily every day. Would be curious to hear if something looks off with that one as well.)

 

magicserver-diagnostics-20240129-2043.zip

[Magic815]

Just to confirm my suspicion, my docker log check today shows that it is indeed growing by 28.3MB every time my container runs (which is daily).

 

 

Any thoughts from anyone? Is there a better spot I should be asking this sort of question?

 

*Edit: I'm going to give this suggestion a try, and then will report back: 

 

Edited January 30 by Magic815

[Magic815]

It looks like the above 'extra params' fixed the issue. And I realized that there is a general setting under the docker service in unraid where one can enable log rotation. I'm considering this issue solved via one of those two methods.