NFS write error on 2650, RH 7.3

David C. Kovar kovar at kealia.com
Tue Mar 18 19:51:00 CST 2003


Good afternoon,

The file server is running 2.4.18-26.7.xsmp. The clients are running one
of 2.4.18-19.7, 2.4.18-26.7, or 2.4.18-27.7. We experience the same NFS
hang at the conclusion of the file write in each case.

-David

On Tue, 2003-03-18 at 00:41, Mansell, Gary wrote:
> What kernel version are you running - I experienced nfs subsytem lockups with 2.4.18-5smp. this was due to a bug in lockd which was fixed in later versions.
> 
> 
> Gary Mansell
> Senior Technical Analyst
> IT Department
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Ricardo Consulting Engineers Ltd.          Direct Line:    01273 794485
> Bridge Works,                              Switchboard:    01273 455611
> Shoreham By Sea,                             Facsimile:    01273 794699
> West Sussex.                                  Internet: www.ricardo.com
> BN43 5FG                                  E-Mail: GRMansell at ricardo.com
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 
> 
> -----Original Message-----
> From: David C. Kovar [mailto:kovar at kealia.com]
> Sent: 18 March 2003 00:02
> To: linux-poweredge at dell.com
> Subject: NFS write error on 2650, RH 7.3
> 
> 
> Good afternoon,
> 
> We're running RH 7.3 with the latest kernel on a 2650. If we try to
> write one specific file out to a NFS partition mounted from a nearly
> identical 2650 fileserver, the write and the entire NFS mount hangs. We
> need to restart NFS on the client to clear the problem.
> 
> The fileserver has ext3 partitions on a hardware RAID. Latest 7.3
> kernel, very generic system.
> 
> I can reproduce the problem reliably using 7.3 clients. 
> 
> 7.1 clients do not exhibit this problem. 
> 
> Writing this file to a nearly identical fileserver that was cloned from
> the original one does not exhibit this problem.
> 
> Writing the file on the fileserver itself does not cause the problem.
> 
> A tcpdump of the end of the exchange produces the following. If anyone
> can assist me in determining what is going on here, I would greatly
> appreciate it. 
> 
> -David
> 
> These will cycle endlessly. And yes, the fileserver's name is
> "fileserver".
> 
> 
> 15:24:48.333528 bee02.2349057636 > fileserver.nfs: 1472 write fh
> Unknown/010000020008000501800000C60279005EA646ABE80279000000000000006000
> 4096 bytes @ 0x000006000 <unstable> (frag 36636:1480 at 0+) (ttl 64, len
> 1500)
> 15:24:51.130036 bee02 > fileserver: (frag 36637:1276 at 2960) (ttl 64, len
> 1296)
> 15:24:51.130049 bee02 > fileserver: (frag 36637:1480 at 1480+) (ttl 64, len
> 1500)
> 15:24:51.130059 bee02.2349057636 > fileserver.nfs: 1472 write fh
> Unknown/010000020008000501800000C60279005EA646ABE80279000000000000006000
> 4096 bytes @ 0x000006000 <unstable> (frag 36637:1480 at 0+) (ttl 64, len
> 1500)
> 15:24:56.723114 bee02 > fileserver: (frag 36638:1276 at 2960) (ttl 64, len
> 1296)
> 15:24:56.723126 bee02 > fileserver: (frag 36638:1480 at 1480+) (ttl 64, len
> 1500)
> 
> _______________________________________________
> Linux-PowerEdge mailing list
> Linux-PowerEdge at dell.com
> http://lists.us.dell.com/mailman/listinfo/linux-poweredge
> Please read the FAQ at http://lists.us.dell.com/faq or search the list archives at http://lists.us.dell.com/htdig/
-- 
David C. Kovar <kovar at kealia.com>




More information about the Linux-PowerEdge mailing list