H
Helpdesk
Hey everyone,
We are runing a Netfinity 5100 dual 933MHz with 1GB mem, RAID5
Win2000 SP4 Exchange 2000 SP3
My server jumps to 80 to 100% CPU Utilization caused by Interrupts/sec
jumping from average 1200 to 70000 (dual CPU)
I need to reboot the server to get it back to normal, for about 48 hours,
and then all over again.
I need help in drilling down the cause of this issue causing high CPU
utilization and slow response on Exchange.
High Interrupts/sec ocurr randomly after maybe 8 to 48 hours of operation.
What i found is that in "normal" operation CPU 0 has about 1100
interrupts/sec and CPU 1 about 150
Question 1. Is this normal CPU behaviour? should it be balanced?
Question 2. Can I savely assume that I have a hardware Issue?
Question 3. What other counter do I need to pull up to isolate the device in
question?
Question 4. Is there a Windows tool that shows me the Kernel times per
device or process?
Any suggestions are welcome.
Jan.
We are runing a Netfinity 5100 dual 933MHz with 1GB mem, RAID5
Win2000 SP4 Exchange 2000 SP3
My server jumps to 80 to 100% CPU Utilization caused by Interrupts/sec
jumping from average 1200 to 70000 (dual CPU)
I need to reboot the server to get it back to normal, for about 48 hours,
and then all over again.
I need help in drilling down the cause of this issue causing high CPU
utilization and slow response on Exchange.
High Interrupts/sec ocurr randomly after maybe 8 to 48 hours of operation.
What i found is that in "normal" operation CPU 0 has about 1100
interrupts/sec and CPU 1 about 150
Question 1. Is this normal CPU behaviour? should it be balanced?
Question 2. Can I savely assume that I have a hardware Issue?
Question 3. What other counter do I need to pull up to isolate the device in
question?
Question 4. Is there a Windows tool that shows me the Kernel times per
device or process?
Any suggestions are welcome.
Jan.