[OpenAFS] Connection timed out?

[John Koyle]

Hi Jim,

I don't suspect this to be the case in our instance.  Only 1 of 7
volumes are affected by this problem.  If it were an mtu problem
wouldn't it affect all the volumes? It seems to be something related to
the volume itself, rather than the client because this happens on 2
separate clients - at nearly the same time.  It's got me puzzled and
afraid to roll this into a production environment.

Thanks,
John

On Thu, 2003-04-03 at 09:26, Jim Rees wrote:
>   Same problem here. Every filetransfer seems to time out after a
>   certain amount of time or bytes.
> 
> I would suspect mtu.  There were certainly bugs in the interface mtu code in
> rx_kcommon.c.  In some cases rxi_FindIfMTU() would just return the first
> interface, which on many systems is the loopback and has an mtu of 32K.  I
> think I fixed this but only time will tell.
> 
> To test this theory, hotwire rx_kcommon.c:rxi_InitPeerParams() so it always
> sets pp->ifMTU = RX_PP_PACKET_SIZE.  Do this only as a test.  If it works,
> your connections won't time out but they won't go very fast either.
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info
-- 
John Koyle
IT Manager
RFP Depot, LLC.