December 2, 20241 yr The cache pools is raid1 with 1tb x 2 ssd. There is only 2 folders in this cache, one store docker.img, one stores libvirt.img /mnt/nvme_1_cache/system/docker/docker.img (825 GB) /mnt/nvme_1_cache/system/libvirt/libvirt.img (8.59 GB) Docker volume usage Total devices 1 FS bytes used 77.33GiB devid 1 size 768.00GiB used 233.02GiB path /dev/loop2 [Problem?] Nvme_1_cache: used 907GB free 91.9GB And free space is decreasing a bit (MB or even GB) day by day (sometimes increase again but not much) even though docker.img and libvirt.img size is not increasing. What am I doing wrong? unraidns-diagnostics-20241202-2153.zip Edited December 2, 20241 yr by Hakun
December 3, 20241 yr Community Expert soulds like a docker is runing and is misconfigured in its storage where it writing to the docker image and not teh disk in the uraid array. I would recomend user script plugin and run these 2 scripts: clean up logs: clean up any and all non running docker: otherwise you will need to capture the logs in a docker that misconfigured.
December 3, 20241 yr Author As my understanding, docker cannot use over its allocated vdisk size. As shown, it now only use 233.02GiB/768.00GiB. About docker cleaning user scripts, I run its several time, nothing change.
December 3, 20241 yr Community Expert Why on earth do you have a docker.img that is 768GiB??? Are you running 300+ docker containers? Go to the docker tab and click "Container Size" at the bottom. What does that output? For example: Edited December 3, 20241 yr by MowMdown
December 4, 20241 yr Author Why on earth do you have a docker.img that is 768GiB??? Are you running 300+ docker containers? <=== I'm an unraid tester ^^
December 4, 20241 yr Community Expert I see your using ai stuff as well. Make sure your models are being saved to the pool/array disk and not the docker image. From ollama, and open-webui. Yes, when testing some dockers especially the AI one due tot here size i have had a 250 GB docker image. Edited December 4, 20241 yr by bmartino1
December 5, 20241 yr Author Quote Make sure your models are being saved to the pool/array disk and not the docker image. The point is docker has not yet used all its allocated size. That is why I don't understand why free cache pool space is being reduced.
December 5, 20241 yr Community Expert Because you have a memory leak or misconfiguration of data within a docker saving to a /tmp location within the docker image. This is hard to isolate pinpoint and fix.
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.