HELP! Oracle 7.3.2 hangs after upgrade

Greetings.

 I hope someone here can help me. I work in a software development environment where we recently upgraded to HP-UX 10.10 and Oracle 7.3.2 Parallel (w/ 2 CPUs). Since, database performance is in the toilet and Oracle 'hangs' more or less at random, but most often when doing something with a large block of data such as imports, deletes, and using sqlload. Our DBAs are working on it; but have not been able to resolve the problem. The "cost" is that my team is working 12-16hrs/day to maintain our schedule.

I have noticed the following symptoms:

• locking shows the activity to have both a table lock and a row lock. typically 3 locks are in place when it hangs, 2 MX and 1 TX.
  • we've cranked the redo log swich rate up to switch every 30 seconds. When deleting a chunk of about 1000 records, it hung and the database was unable to perfrom a checkpoint and unable to switch the redo log file. killing the delete and restarting it in much smaller increments seemed to resolve the problem.
  • some of the hangs are so bad that the system administrator has to kill the job.

It looks to me like the process is blocking a system activity and nothing is catching it. A recent change in semaphores perhaps. I not an expert in Unix or Oracle (although I play one at work <g>) and wonder if anyone has experienced this problem and/or has a clue what is going on.

Thanks in advance. Received on Fri Dec 12 1997 - 00:00:00 CET