[Bug 260664] ZFS/NFS: Intermittent hangs and crashes after a period of time in: nfscl_hasexpired || dbuf_write_done || zio_execute
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Sat, 25 Dec 2021 00:27:24 UTC
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=260664 --- Comment #2 from Rick Macklem <rmacklem@FreeBSD.org> --- I'll make a few random comments. 1 - Since the crashes occur at different places, I agree with you that it might be a hardware problem. It would be nice if you could set up another machine with exactly the same software/usage and see if it crashes as well. 2 - You should never use "intr" nor "soft" on NFSv4 mounts. This is mentioned in the BUGS section at the end of "man mount_nfs". To get rid of a hung NFS mount, use "umount -N <mnt_path>. (It can take a couple of minutes, but normally succeeds. Note that any file writing that was in progress when you do this will get lost.) 3 - If you can still log in and do these when "hung", capture the output of: # ps axHl # procstat -a -kk # netstat -a # ping <nfs-server> 4 - nfscl_hasexpired() only gets called when the NFS client does not get a response from the server for minutes and then gets a NFSERR_EXPIRED reply from the server. You did not mention what NFS server you are using. If it is a FreeBSD 13.0 server, then see PR#256280. If you get the output from #3, please post it here. -- You are receiving this mail because: You are on the CC list for the bug. You are the assignee for the bug.