On Thu, 11 Sep 2003 10:12:33 -0400, Donovan O'Brien <dobrien AT SIAC DOT COM>
wrote:

>We had the same annoying problem for almost a year.  Open call
>with Legato finally discovered the problem.  We had isolated this
>to a few clients but could not seem to find out what was hanging
>the drives.  Clients would stream with a good speed and then
>abruptly halt.  Drive would then become useless and all save sets
>from other clients writing to that drive would also fail.  We
>found the error to be a tcp setting on the HPUX client.  Seems like
>the users had a program that used ndd to set the tcp_keepalive_interval
>to 60 seconds from the default of 2 hours.  The save
>sessions would time out and close the tcp sessions.  Once the parameter
>was reset, we have never seen the condition again.  I suggest you
>take a look and see if a particular client or set of clients are
>in the savegroup that is causing this condition and troubleshoot
>from there.  Good luck.

There could be other reasons for the tcp_keepalive_interval being set that
low but one I've heard of before is to get round a firewall initial timeout
problem - of course that might not be issue for you.

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=