RHEL4 U3 with 2 Qlogic 2342 HBAs on sunv40z
Hello,
One of our performance teams is experiencing lockups when they try to run light sequential workloads with 2 Qlogic 2342 Fibre Channel HBAs on the same PCI bridge.
During these runs, they are also seeing this message quite frequently:
Aug 24 07:16:44 v40z2 kernel: warning: many lost ticks.
Aug 24 07:16:44 v40z2 kernel: Your time source seems to be instable or some driver is hogging interupts
Aug 24 07:16:44 v40z2 kernel: rip __do_softirq+0x4d/0xd0
I suspected that this might be due to an irq conflict, and suggested that they move one of the boards to a different slot. This seems to have improved the situation. However, it would be good to know whether or not this configuration is supported, etc.
Any information you can provide would be very helpful.
Thanks a lot,
Claire
