On 03/29/2010 12:30 AM, Colin Brace wrote:
Hi all,
I have a small LAN server running FreeBSD and a desktop with F12 installed. Both systems have several shares available via NFS, which I mount on a laptop, also running F12.
This setup as served me well for three or four years.
Since a certain point last fall, beginning with F11, I have been having recurring problems writing from the laptop to the NFS shares. Although they are mounted rw, the moment I try to copy something to a remote share (either in Nautilus or in a terminal), my mounted shares die and Nautilus freezes up. I try restarting NFS, killing and restarting nautilus, to no avail; my only recourse is to reboot.
The export files on my F12 desktop and Freebsd look like this:
F12: /media/disk1 192.168.2.0/24(rw,sync,nohide) /media/disk2 192.168.2.0/24(rw,sync,nohide) /media/disk3 192.168.2.0/24(rw,sync,nohide) /media/disk4 192.168.2.0/24(rw,sync,nohide) /media/disk5 192.168.2.0/24(rw,sync,nohide) /media/disk6 192.168.2.0/24(rw,sync,nohide)
FreeBSD: /usr -alldirs -network 192.168.2.0/24 /media/disk7 -network 192.168.2.0/24 /media/disk8 -network 192.168.2.0/24
These are the entries in fstab on the F12 client:
ariel:/media/disk1 /media/disk1 nfs4 defaults,auto,users,rw 0 0 ariel:/media/disk2 /media/disk2 nfs4 defaults,auto,users,rw 0 0 ariel:/media/disk3 /media/disk3 nfs4 defaults,auto,users,rw 0 0 ariel:/media/disk4 /media/disk4 nfs4 defaults,auto,users,rw 0 0 ariel:/media/disk5 /media/disk5 nfs4 defaults,auto,users,rw 0 0 ariel:/media/disk6 /media/disk6 nfs4 defaults,auto,users,rw 0 0
venus:/media/disk8 /media/disk8 nfs rw,soft,bg,rsize=32768,wsize=32768,nfsvers=3,tcp,intr 0 0 venus:/media/disk7 /media/disk7 nfs rw,soft,bg,rsize=32768,wsize=32768,nfsvers=3,tcp,intr 0 0 venus:/usr/home/colin /media/venus nfs rw,soft,bg,rsize=32768,wsize=32768,nfsvers=3,tcp,intr 0 0
There is some ugly stuff in the message log; don't know what it means:
Mar 28 22:03:12 hera mountd[1516]: Caught signal 15, un-registering and exiting. Mar 28 22:03:12 hera kernel: nfsd: last server has exited, flushing export cache Mar 28 22:03:13 hera kernel: NFSD: Using /var/lib/nfs/v4recovery as the NFSv4 state recovery directory Mar 28 22:03:13 hera kernel: NFSD: starting 90-second grace period Mar 28 22:03:46 hera kernel: INFO: task nautilus:2967 blocked for more than 120 seconds. Mar 28 22:03:46 hera kernel: "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. Mar 28 22:03:46 hera kernel: nautilus D 000001b3 0 2967 1740 0x00000080 Mar 28 22:03:46 hera kernel: e92d1e6c 00200082 b231c8c5 000001b3 0000330d 00000000 eda10f6c 00200246 Mar 28 22:03:46 hera kernel: c0a5d6ac c0a62120 eda10f6c c0a62120 c0a62120 e92d1e5c c045b6bb 4d8238bb Mar 28 22:03:46 hera kernel: 00000001 eb6b7c00 000001b3 eda10cc0 c2a08120 00000000 e92d1eb4 e92d1e64 Mar 28 22:03:46 hera kernel: Call Trace: Mar 28 22:03:46 hera kernel: [<c045b6bb>] ? ktime_get_ts+0x98/0xa2 Mar 28 22:03:46 hera kernel: [<c0791937>] io_schedule+0x37/0x4e Mar 28 22:03:46 hera kernel: [<c04a54a2>] sync_page+0x3f/0x43 Mar 28 22:03:46 hera kernel: [<c0791df6>] __wait_on_bit+0x39/0x60 Mar 28 22:03:46 hera kernel: [<c04a5463>] ? sync_page+0x0/0x43 Mar 28 22:03:46 hera kernel: [<c04a56a1>] wait_on_page_bit+0x80/0x89 Mar 28 22:03:46 hera kernel: [<c0454229>] ? wake_bit_function+0x0/0x3c Mar 28 22:03:46 hera kernel: [<c04a5e36>] wait_on_page_writeback_range+0x52/0xf0 Mar 28 22:03:46 hera kernel: [<c04a5ff9>] filemap_write_and_wait_range+0x4e/0x5d Mar 28 22:03:46 hera kernel: [<c04f6a8c>] vfs_fsync_range+0x4e/0x91 Mar 28 22:03:46 hera kernel: [<c0480795>] ? audit_syscall_entry+0x116/0x138 Mar 28 22:03:46 hera kernel: [<c04f6b46>] vfs_fsync+0x19/0x1b Mar 28 22:03:46 hera kernel: [<c04f6b6f>] do_fsync+0x27/0x37 Mar 28 22:03:46 hera kernel: [<c04f6ba8>] sys_fsync+0x12/0x14 Mar 28 22:03:46 hera kernel: [<c040365c>] syscall_call+0x7/0xb Mar 28 22:03:46 hera kernel: INFO: task evince-thumbnai:2978 blocked for more than 120 seconds. Mar 28 22:03:46 hera kernel: "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
What I do notice is that when I run 'service nfs restart', "nfs services" is always dead and never seems to restart after this point. Looking at the init.d file, it appears exportfs is hosed in some way.
I have been tinking with the export files and Googling this problem endlessly without finding a solution, and I am hoping someone here might have some ideas on what is going wrong here.
Thanks.
Colin Brace Amsterdam http://lim.nl
Does this occur with an older kernel?