nfs client hanging

[Guillaume Rousse]

Trond Myklebust a écrit :
> On Tue, 2008-06-17 at 16:01 +0200, Guillaume Rousse wrote:
>> Trond Myklebust a écrit :
>>> However the fact that everything is being queued on the xprt_backlog
>>> means that there is a heavy congestion, and that all the RPC slots are
>>> full. It supports the suspicion that the server is failing to respond to
>>> the client, and so the client requests are all backed up.
>>>
>>> Does 'netstat -t' show the client as still being connected to the
>>> server?
>> I had to reboot the server, unfortunatly, I can't answer :(
>>
>> However, I regulary have this in he logs, even right now after the reboot:
>> Jun 17 14:17:13 chatelet kernel: RPC: failed to contact local rpcbind
>> server (errno 5).
>>
>> Despite rpcbind process currently running. With rpcdebug flags call & 
>> bind (I choose those ones because I had no better clue, and all is a bit 
>> too much verbose) activated, I have this:
>>
>> Jun 17 14:27:28 chatelet kernel: RPC: failed to contact local rpcbind 
>> server (errno 5).
>> Jun 17 14:27:28 chatelet kernel: RPC:       registration status 
>> -5/-532611276
>>
>> is this related ?
> 
> No: the kernel will be talking to the rpcbind server on a completely
> different socket, and normally it should not be using rpcbind/portmap to
> figure out which port the filer's NFS server is using.
> 
> That said, have you set CONFIG_SUNRPC_BIND34 in your .config file?
No. And IPv6 is disabled anyway on those hosts.
-- 
Guillaume Rousse
Moyens Informatiques - INRIA Futurs
Tel: 01 69 35 69 62