In an attempt to figure out what is going wrong on these machines, on
box2+3 I have unchecked "used random port assignments" - I will see if
this fixes the backups tonight, thanks for all the help/ideas/etc!



-----Original Message-----
From: veritas-bu-admin AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu] On Behalf Of Paul
Keating
Sent: Tuesday, November 08, 2005 1:11 PM
To: veritas-bu AT mailman.eng.auburn DOT edu
Subject: FW: [Veritas-bu] Very ODD problem with NB5.1MP3a + 2 Solaris
Clients


-----Original Message-----
From: Paul Keating 
Sent: November 8, 2005 1:02 PM
To: veritas-bu AT mailman.eng.auburn DOT edu
Subject: RE: [Veritas-bu] Very ODD problem with NB5.1MP3a + 2 Solaris
Clients


yep, completely different subnets and routes, but for some reason, I'd
bet that when you try to do a backup, the client is trying to reply to
the server on the wrong interface.

I'll bounce it off my network guys.....I can't remember the details on
their side...I think one of the guys even did a powerpoint presentation
on it diagramming the traffic flows.

it is in essence a box problem...im my case, it was with a windows file
server, and an exchange server.

In both cases, there was "only one path to the backup server, and other
NICs but with separate interfaces+routes+subnets"


you showed that after you hit the backup server from the client, THEN
you could hit the client from the server....but for how long?
5 min? 10 min? 30 min?

figure that out, then ask your network guys what the ARP refresh rate
and CAM table timeouts are on the switches/routers.

this only happens with 2 of the 5 clients, correct? are they connected
to a different switch, or is their "non-backup vlan" different than the
ones that do work?

Paul