As part of our performance monitoring, I track certain processes using PS.
By taking two snapshots and comparing the CPU usage, it is possible to
identify the BIG CPU users for long running processes.
I have noticed the following problem though.
An Oracle Checkpoint process ora_abc_ckpt, managed to change its name to
[oracle] for the second PS. I ran PS again and the name had changed back to
ora_abc_ckpt. The pid remained the same, and the CPU count increased at
each measurement.
My questions are basically:
a) Why does this happen?
b) Should I be concerned?
c) How can I stop this from happening?
For further information, the only process name that I can find that
consistently uses [...] is [kernel idle]. I know what this process does, so
there is no need to explain this, but I am wondering if the use of [...] in
a process name indicates a specific kind of process state.
Your help is appreciated
Received on Fri Sep 03 1999 - 14:10:54 NZST