I utilize podman on my server for running my software.
Recently, I saw that my server ran out of space on its 8TB raid 10 array. Which immediately raised questions for me.
After using “duc” to analyze my drive. I found that podman had used 4 TB of storage space in “/var/tmp”.
Anyone else have this happen to them?
In my case I like to run most things in containers, and that includes a lot of ML shit. Since I’m on AMD nothing fucking works, so I accumulate all these huge multiple GB containers that don’t work because the whole ML/AI/GPU stack is a steaming pile.