Priority Checking - For database Hanging

Assumptions :The database is running in one of the unix platform.

Check for locks using standrad scripts as described below:

1. Check if numerous locks exist in the database(You can use some scripts _ I will try to put in this blog soon)
2. Look for TX locks/ Kill the processes which are form / CM long running sessions/ Run -away processes.
3. If SS- Enqueue/ ST - Enqueue , then we need to dig further , which process is creating
4. If oracle background processes , then try to find whcih one.
2. Check for the services running in that node and may consider moving some of the services from that node to other nodes which are a bit lesser burdened.

3. Do a top -h for looking which process (oracle/ user) is taking much cpu.
$ top -h
or
$ glance

4. Open a SR with oracle and upload the following files:
1. alertlog
2.Hang Analyze (means at node level)
3.CRS oswatcher files
4. Systemstate Dumps
5.Processstack
6.Errorstack at level 3 / or at level 1
7.Hang analyze Global
8.Processstack for the background processes for oracle background processes.
9.AWR report ( Execute $ORACLE_HOME/rdbms/admin/awrrpt.sql) -In 10g its awr , in 9i - it was statspack)

I will discuss more on each with individual sessions in coming days.