UNIX for Dummies Questions & Answers

Hi,

I currently have two processes running on my Solaris server that are hogging the cpu. I am confident that these two processes are caused by remote users logged into the oracle database on the server.

What I need to know is how to get the corresponding ip address associated with the process ID's of the two processes. I'm guessing the netstat command must be involved here? Thanks for your help.

With netstat you can see on which socket/tcp ip port they are connected, but not which process actually. I don't have Solaris so I can't check, but on Debian Linux or AIX I don't know of listing associated PIDs to connections.
Sometimes the tool lsof is very helpful in identifying such stuff.

Did you see with ps that oracle processes are pushing the CPU(s) to the limit?
To get an overall performance view of the machine I would recommend vmstat.

From DB2 I remember that there are some tools to check traffic in view of the RDBMs - I bet there is something similar to that on Oracle.

Thanks Zaxxon.

I do know about the lsof command but unfortunately it is not installed on this server. Nor will I be allowed to install it without a moutain of paperwork. I used vmstat to monitor the system and the fundamentals seem to be ok. I'm using prstat 1 to look at the processes which are hogging cpu.

I will ask the oracle DBA's about monitoring tools and if they have used them before. I was just looking for an ip address so I could then get a username and ask them if they were running anything that required a lot of processing power. The system is back to normal now as we found the culprits. Thanks for your reply.