N5200 NFS stopped

[asmuir]

The NFS daemon on our N5200 has stopped and can't be restarted.
When I first reboot NFS starts and then dies. I logged in with
SSH and typing dmesg for the system log which shows that the nfsd
has crashed with a trace message.
The problem first started while running 2.004 (it worked before on 2.004)
I then upgraded to 2.008 but it made no difference.

Any ideas?

[Thecus - Yvon]

Dear Sir,

Please update to latest version V2.00.10 to verify NFS function. If still not working and need Thecus ESUPPORT assist, please reset to facotry default to rid off SSH affect and try the NFS again, if nessary please provde the CONF.BIN for us analysis for you, from http://esupport.thecus.com/


[N5200/N5200PRO]
Latest firmware: v2.00.10
Release note: http://www.thecus.com/Downloads/N520...WRN_v20010.pdf
Upgrade procedure: http://www.thecus.com/Downloads/N520...UPG_v20010.pdf
Download link: http://www.thecus.com/Downloads/N520...FW_2.00.10.bin
FAQ: http://www.thecus.com/download/other...2006-11-22.pdf

[pit234a]

There was an bug in 2.00.04, not described anywhere, but I am sure about that. NFS wouldn't work correct as it should according to the documentation. It once in a while run accidentally, but to get it work properly, I needed help and some extra modules. It seems, that this error is fixed with 2.00.10 now, but I still operate my 2.00.04 with those modifications. I would want those modules anyway, SYSUSER, SSHD, NFSEXPORT, UTILS and META, to name them. So, no reason for me to switch over to a newer FW.

As far as on N5200, you may check this:

root@127.0.0.1:/raid0/data# ps | egrep "nfs|mount|lock|stat|rpc|port"
    7 root            SW< [kblockd/0]
 2498 1           452 S   /sbin/portmap
 2778 root        368 S   /opt/nfs-utils/sbin/rpc.rquotad
 2780 root            SW  [nfsd]
 2781 root            SW  [nfsd]
 2782 root            SW  [nfsd]
 2783 root            SW  [nfsd]
 2784 root            SW  [nfsd]
 2785 root            SW  [nfsd]
 2786 root            SW  [nfsd]
 2787 root            SW  [nfsd]
 2788 root            SW  [lockd]
 2789 root            SW< [rpciod/0]
 2791 root        812 S   /opt/nfs-utils/sbin/rpc.mountd
 2804 root        464 S   /opt/nfs-utils/sbin/rpc.idmapd

From your PC (given you use some kind of Unix on it, what you better should):

pit@syo ~:-> showmount -e 192.168.0.100
Exports list on 192.168.0.100:
/raid0/data/pit                    *.*.*.*
/raid0/data/usbcopy                *
... and more ...

pit@syo ~:-> nmap 192.168.0.100

Starting Nmap 4.68 ( http://nmap.org ) at 2008-08-14 11:14 CEST
Interesting ports on 192.168.0.100:
Not shown: 1705 closed ports
PORT     STATE SERVICE
21/tcp   open  ftp
22/tcp   open  ssh
80/tcp   open  http
111/tcp  open  rpcbind
443/tcp  open  https
548/tcp  open  afp
631/tcp  open  ipp
850/tcp  open  unknown
2000/tcp open  callbook
2049/tcp open  nfs

Nmap done: 1 IP address (1 host up) scanned in 0.297 seconds

nmap should exists for M$ also.

[asmuir]

Thanks for your replies. An update:

I have now upgraded to 2.00.10 as recommended. However this makes no difference
to the problem - NFS still stops immediately after boot up.

Note that NFS was working fine on 2.00.04 for many months, but suddenly stopped
a few days ago. No modules had been installed at all - I only installed SSH to have
a look at what was happening after upgrading to 2.00.08, so I don't think it has anything
to do with any 3rd party modules.

the dmesg log I mentioned shows the nfsd crashing after bootup:

CPU: 0
EIP: 0060:[<c0161599>] Not tainted VLI
EFLAGS: 00010246 (2.6.13N5200)
EIP is at permission+0x49/0xa0
eax: ddd27800 ebx: 00000001 ecx: ddd27800 edx: 00000000
esi: c7a9984c edi: cee1efc0 ebp: c7a9984c esp: cb09bd24
ds: 007b es: 007b ss: 0068
Process nfsd (pid: 4722, threadinfo=cb09a000 task=cb227ac0)
Stack: c80c1164 00000000 c0162cde c7a9995b d048312d ddd27800 00000001 00000000
c7e17a6c c80c1164 c7e17a6c d01fb2ae cee1efc0 c7a9984c 00000008 00000000
c0350450 00000000 c7a9984c 73726570 6c616e6f 6c6f2e00 75b60064 00000000
Call Trace:
[<c0162cde>] lookup_one_len+0x6e/0x80
[<d048312d>] nfsd_acceptable+0x6d/0xf0 [nfsd]
[<d01fb2ae>] find_exported_dentry+0x2ae/0x5b0 [exportfs]
[<c0117005>] activate_task+0x65/0x80
[<c0117981>] __wake_up_common+0x41/0x70
[<d04892a5>] exp_find_key+0x85/0xa0 [nfsd]
[<d01fb93a>] export_decode_fh+0x5a/0x7a [exportfs]
[<d04830c0>] nfsd_acceptable+0x0/0xf0 [nfsd]
[<d04833bb>] fh_verify+0x20b/0x5b0 [nfsd]
[<d04830c0>] nfsd_acceptable+0x0/0xf0 [nfsd]
[<c0117005>] activate_task+0x65/0x80
[<d0484adf>] nfsd_access+0x2f/0x100 [nfsd]
[<d048cfb8>] nfsacld_proc_access+0x88/0xc0 [nfsd]
[<d0481607>] nfsd_dispatch+0xd7/0x200 [nfsd]
[<d0481530>] nfsd_dispatch+0x0/0x200 [nfsd]
[<d044a790>] svc_process+0x550/0x690 [sunrpc]
[<d04813bd>] nfsd+0x19d/0x310 [nfsd]
[<d0481220>] nfsd+0x0/0x310 [nfsd]
[<c0101ded>] kernel_thread_helper+0x5/0xb
Code: f6 40 34 01 74 1c 89 d0 25 00 f0 00 00 3d 00 80 00 00 74 60 3d 00 40 00 00 74 59 3d 00 a0 00 00 74 52 f6 81 14 01 00 00 08 75 42 <8b> 91 8c 00 00 00 83 e3 f7 85 d2 74 07 8b 42 34 85 c0 75 1a 89
<6>md: md0: sync done.
criticalevent_user: raid_degrade md0
RAID1 conf printout:
--- wd:4 rd:5
disk 0, wo:0, o:1, dev:sda1
disk 1, wo:0, o:1, dev:sdb1
disk 2, wo:0, o:1, dev:sdc1
disk 3, wo:0, o:1, dev:sdd1
Read after write, bit[35]=1
scsi: unknown opcode 0x0c

Also:

root@127.0.0.1:~# ps | egrep "nfs|mount|lock|stat|rpc|port"
7 root SW< [kblockd/0]
4727 root SW [lockd]
4728 root SW< [rpciod/0]
7028 1 512 S /sbin/portmap
7605 root 684 S /opt/nfs-utils/sbin/rpc.rquotad
7608 root 888 S /opt/nfs-utils/sbin/rpc.mountd
7610 root 708 S /opt/nfs-utils/sbin/rpc.statd
7623 root 1020 S /opt/nfs-utils/sbin/rpc.idmapd

Thanks

[pit234a]

First, this looks like a kernel problem to my eyes, but with changing FW, it should than be gone. Therefore, there is not a lot whats left. I assume, nfsd dies, because the wanted export is invalid. One hint to that might be my line in dmesg:
md: md0: sync done.
criticalevent_user: raid_healthy md0
raid_healthy sounds better to me than raid_degrade in your log.
Can you confirm, that your raid is ok?
Maybe, you give it a try to not export anything from the raid, but, maybe from system or maybe from some other drive and see what happens than.
(/app/bin/rc/nfs restart)&
or start instead restart should start nfsd and (/app/bin/rc/portmap restart;/app/bin/rc/nfs restart)& will restart and read exports again and /opt/nfs-utils/sbin/exportfs -r only rereads the exports and uses this.
Maybe, you get more helpful information with that.