On 14/04/18 08:17, Tony Lewis wrote: > I am only just starting with aufs, and my context is using it to overlay > eight disks to form a kind-of RAID. > > I find three symptoms: > > 1. If I do a 'find' command on the aufs-mounted directory (e.g. 'find > /mnt/merge/data -print') it starts fine. If I do Ctrl-C to terminate, > it doesn't. I cannot kill the find command even with kill -9 as root > > 2. If I have used the mounted directory, then the server will not shut > down completely. The console has warnings like 'INFO: task > umount.aufs:3746 blocked for more than 120 seconds.' > > 3. If I do a df -h then the reported free space and available space is > for the first branch added, and not for the overall disk. > > My system is a near-pristine Debian stretch, AMD64. The disks are 1.0TB > SATA, and seem to be OK (no errors other than non-serious ones in > smartctl), though they are ageing. > > Kernel is 4.9.0-6-amd64, as per Debian repos. aufs-tools is > 1:4.1+20161219-1, which I understand to be based on aufs v4.1, c. > 2016-12-19. > > Things I've tried: > > * unmounting before shutting down: this then shuts down cleanly (but > doesn't solve the 'find' problem > > * mounting only a single branch: this still causes 'find' to be > unkillable and shutdowns to be incomplete > > Can anyone offer suggestions of what to try next? > > Tony
Possibly not directly helpful, but I've had many instances where aufs (4.9-20171218 and earlier) locks up a process in 100% kernel-mode processing (usually associated with find/locate), requiring hardbooting to escape from. For a few months I've been rebuilding aufs on every new kernel package, new kernel version or not, and haven't had the problem since.
signature.asc
Description: OpenPGP digital signature
------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
