telnet connections broken

[Hr. H. Wichink Kruit]

We've have a RP7405 server with HPUX 11.0.

we also have a wan devided over 3 locations and connected by a wireless connection and Cisco routers.

The PC's in the other locations are connected by the HP UX server with a telnet connection.

Problem.
Regulary the wireless connection is broken and a ISDN backup line will be activated.
This consumes a little time (about 15 pings).
But this is long anough to disconnect the telnet sessions resulting in broken database logins.

Is it possible to set a few parameters on the Unix server to preserve these telnet sessions longer. I am thinking of a kind of keep alive interval.
I know this is very brief but maybe someone recognises the problem and can help me.

H. Wichink Kruit

[Ron Kinner]

I expect you are getting shutdown by:

tcp_ip_notify_interval:

First threshold timer for established connections.

If the first threshold is exceeded, TCP notifies IP that it
is having trouble with the current established connections
and requests IP to delete the routing table entry for this
destination. The assumption is that if no ACK has been
received for an extended period of time, there may be
network routing problems and IP should try to find a new
route. [500,-] Default: 10000 (10 seconds)

You can try changing it to a value slightly greater than the time it takes to setup the ISDN and see if that helps:

ndd -set /dev/tcp tcp_ip_notify_interval 35000

would set it to 35 seconds which I would expect would be about 5 seconds more than 15 ping timeouts.

However, the other end may need to have the same change made.

If it does help then you need to edit /etc/rc.config.d/nddconf to add

TRANSPORT_NAME[0]=tcp
NDD_NAME[0]=tcp_ip_notify_interval
NDD_VALUE[0]=35000

If you already have entries in nddconf then use the next higher digit for the number in brackets. This will insure that the changes stays after a reboot.

Ron