Watchdog error

[mennoslaats]

I get every 50 hours the next error message:

hptx: Watchdog ReInit Memory 6 board 0 hptx - Warning: hptx: Allocb failure in ReInitMemory

I updated the drivers of my networkcard, without problems. The error message is still there.

Anyone an idea? Is it realy a problem with the Memory or netwerok card?

 

[Annihilannic]

We used to get similar errors with the IntelEtherExpress (driver eeE), but a driver update fixed that.

Annihilannic.

 

[mennoslaats]

When we update the drivers, the error message went away, but the problem with a freezing system system is still here.

 

[rajeshbahl]

Many times this problem is related to high volume of the traffic on the network which your "Hub" can not handle.
I hope your server keeps working fine even when the "freeze" happens at client systems.

Just upgrade your "hub" to "Switch". That should solve this bottleneck.

I have recently done this at one site successfully.


Best Of Luck !
Rajesh K. Bahl

 

[mennoslaats]

We log in every 48 hours to restart the server. So it can't crash after 50 hours.

But we will check the switch.

 

[dpellegrini]

I'm running into a similar problem. We're running a SCO box on a Dell server with an Intel Pro 100 card. The drivers would not update for the NIC card. the Kernel would not recompile. I've got an identical system which is on the same switch (3Com 3300). The ports on the switch are not producing any errors. I'm thinking that replacing the NIC may be necessary. SNMP is enabled on both servers, and I'm wondering if that has anything to do with it. Anyone have any other suggestions?
Thanks in advance.


Domenick Pellegrini
dpellegrini@yahoo.com

 

[vistech1]

This may help.

I explaines that the NSTRPAGES parameter in the kernel may need to be set higher.

http://stage.caldera.com/cgi-bin/ssl_reference?101596