Hi Neil,
Now you have encountered two issue: one is unresponsiveness of card under heavy load, another is crashes under heavy load. Do you have the call stack from crash? There is a fix for crash here for your reference: http://bugs.opensolaris.org/view_bug.do?bug_id=6874776 Regards, Crisson On Sun, 2010-01-17 at 09:35 -0800, Neil McCarthy wrote: > I have built a storage server based on snv_111b on HCL hardware by Dell. > > I have one NIC interface for management and one for iSCSI and NFS (storage) > access. > > Under very heavy load, the storage interface stops responding, the management > card remains unaffected. > > I connect via the management interface and can see that the card is still up. > A snoop reveals that the storage interface is sending data, but appearing to > receive nothing back (I am at a remote location, so have been unable to do a > packet trace on the physical switch to determine if the traffic is actually > being sent to the server, just not being passed along the stack). > > The symptoms seem to be very similar to bug id > http://bugs.opensolaris.org/view_bug.do?bug_id=6727837 - however the > workaround seems to apply to e1000g cards. > > Can anyone offer any advice, as I am getting one to two crashes a week at the > moment, and it is becoming very frustrating. > > Thanks > > Neil _______________________________________________ networking-discuss mailing list networking-discuss@opensolaris.org
