Subject: NFS retransmit timers
To: None <netbsd-users@netbsd.org>
From: Aaron J. Grier <agrier@poofygoof.com>
List: netbsd-users
Date: 07/26/2006 23:36:33
both client and server running GENERIC netbsd-3.0_STABLE.

Jul 26 23:00:02 boromir /netbsd: nfs server server:/disk/media: not responding
Jul 26 23:00:07 boromir /netbsd: nfs server server:/disk/media: not responding
Jul 26 23:16:29 boromir /netbsd: nfs server server:/usr/projects: not responding
Jul 26 23:20:19 boromir /netbsd: nfs server server:/usr/projects: is alive again
Jul 26 23:20:19 boromir /netbsd: nfs server server:/disk/media: is alive again Jul 26 23:20:19 boromir /netbsd: nfs server server:/disk/media: is alive again 

note the times...  server never actually went down; it ran out of
transmit clusters and dropped a few packets.

20 minutes seems excessive.  the retries counter for nfsstat -c on the
client is increasing, but nothing shows up for nfsstat -s on the server.

TCP made things worse in the past, so I went back to UDP.

-- 
  Aaron J. Grier | "Not your ordinary poofy goof." | agrier@poofygoof.com
              "silly brewer, saaz are for pils!"  --  virt