August 6, 20241 yr Requesting IO tuning options for btrfs scrub in the UI (scheduler), or some sensible defaults for file system scrubs. Currently my experience is that the disk(s) become unusable for even regular read operations during scrub, pool size of ~100TB means scrubs takes 2-3 days. relevant scrub options for adoption -c <ioprio_class> set IO priority class (see ionice(1) manpage) -n <ioprio_classdata> set IO priority classdata (see ionice(1) manpage) https://man7.org/linux/man-pages/man8/btrfs-scrub.8.html
August 6, 20241 yr Solution Btrfs scrub should already be low priority by default: Can you detail your pool layout so that I can run some tests?
August 7, 20241 yr Author 6 hours ago, JorgeB said: Btrfs scrub should already be low priority by default: Can you detail your pool layout so that I can run some tests? Thanks Jorge. Makes sense that its actually default, I assumed because it wasn't run with the priority flags it wasn't being deprioritised (or there were lower levels available)... I missed the man note. I have revisited my testing, and reducing the amount of pools scrubbing in parallel has improved the situation, so I think I was being over zealous and overloaded the CPU, or possibly the HBA, both of which are avoidable situations
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.