what is the cause for the Enqueue wait spikes

[sambavan]

I have acqired an instance running 8.7.2 SE on Solaris 8.

How further could I find a the cause for the enqueue waits that spikes up during certain period of times. The time is consistent. I check the locks and the locks seems to be held on one perticular table with row_exclusive. This table has millions of rows and it is all indexed and analyzed.

A further drill down shows inserts and I wonder whether increasing the free list will solve the problem. I have enough dml_locks and enqueue_resource (262000, 270000). What else could I tune to aliviate this contention?

My brains are not thinking right, so thougth of putting this question to the ones who acan think at this moment.

Any suggestions?

Sam

[jmodic]

It is possible you are suffering from ITL contention. Try increasing INITRANS for that table.