eth0: memory shortage

Configuration is as follows:
Tyan 2466 motherboard, 3c905c onboard nic, version
LK1.1.18ac of the 3c59x driver in 2.4.21-20smp version of kernel
as supplied by red hat, on enterprise linux 3.

In previous versions of the OS we also saw

eth0: too much work in interrupt

errors, but now we are seeing them simultaneously across all
240 nodes of a cluster.  Sometimes we also see

eth0:memory shortage

error as well.

Any idea what might be causing this?  Certainly corelated to network
activity in general, particularly if we have nodes rebooting and
doing a lot of nfs activity on the way up and down.  Any reason
why it would be worse on this version of the driver/OS? Any difference
in multicast handling?


Steve Timm

------------------------------------------------------------------
Steven C. Timm, Ph.D  (630) 840-8525  [email protected]  http://home.fnal.gov/~timm/
Fermilab Computing Div/Core Support Services Dept./Scientific Computing Section
Assistant Group Leader, Farms and Clustered Systems Group
Lead of Computing Farms Team
_______________________________________________
vortex mailing list
[email protected]
http://www.scyld.com/mailman/listinfo/vortex