$ svrmgrl Oracle Server Manager Release 188.8.131.52.0 - Production Copyright (c) Oracle Corporation 1994, 1995. All rights reserved. Oracle7 Server Release 184.108.40.206.0 with the 64-bit option - Production Release With the distributed and parallel query options PL/SQL Release 220.127.116.11.0 - Production SVRMGR> connect internal Connected. SVRMGR> @?/rdbms/admin/utlbstat wait some period of time, run your problem queries you want to tune, etc SVRMGR> @?/rdbms/admin/utlestat SVRMGR> exit
The file report.txt contains several sections in this order (I think the order is different for different versions, and I think v$waitstats was added in 7.3)
With the resulting file of some several hundred statistics, the big question is ...
Well, no(w) sweat. The is a short section that called the wait events that will quickly point out tuning problem areas and eliminate other areas as no problems at all.
After having verified wait events, it is also worth while to look at a few statistics if you haven't already done so while interpreting the wait events.
and thats it, more or less. There is little need to look at the other sections of report.txt if the wait events section and statistics section did not point out any problems in those areas.