Sounds like a job for Dtrace. You should be able to use the
fbt provider to find out where the packets go and who in the
kernel drops them. Although my bet would be on IP filtering,
since all of those packets are handled in the IP kernel
module itself, and it seems unlikely that something could
block only one or two type of traffic like that within
the module. Same for the hardware you mentioned, the only
difference between a TCP packet and a UDP packet is a couple
of bits in the header, so it couldn't be a bad cable for
instance.

Martin Gasthuber wrote:
> Hi,
> 
>   one of our X4500 running snv79 stops sending/receiving all other packets 
> except TCP every few 10 minutes for about 10 to 20 minutes. Live TCP sessions 
> (i.e. ssh login) continue to work, but all DNS + ping +... hangs (from in to 
> out and vice versa). We checked this behaviour with an explicit port monitor 
> on the switch side (Cisco). Meanwhile i stopped nearly all processes/daemons 
> not really needed, the cabling from switch to host has been testet/replaced, 
> the switchport ha been changed, we upgrage from snv55 to snv79 - all this 
> makes no change ! Any other idea what we can look for ???
> 
> Thanks,
>    Martin
>  
>  
> This message posted from opensolaris.org
> _______________________________________________
> networking-discuss mailing list
> [email protected]

-- 
blu

"Mr. Jefferson, build up that wall!"
----------------------------------------------------------------------
Brian Utterback - Solaris RPE, Sun Microsystems, Inc.
Ph:877-259-7345, Em:brian.utterback-at-ess-you-enn-dot-kom
_______________________________________________
networking-discuss mailing list
[email protected]

Reply via email to