============================
1.concurrent:program -->define --> enable trace.
2.Now save and run the request. And see the udump file.
3.Now take that and put tkprof by
4.tkprof sample01.trc output=sample.txt
5.Now view the text file by vi editor and in linux, sort the file output by the following command
6.Grep -I total Tkprof outpfile name sort -n +5
7.Now leave the last two consolidated total and look the above highest four count from bottom.
8.Now check the sql which takes the more usage.
====================================================================
possiblities to check and Tune concurrent request:
====================================
1.Go to Concurrent --> Administer -->
2. Now check the the standard manager's Actual and target.
3. Let us assume it is Actual 30 and Target 30.
4. Now check the pending it can be 20
5.check the running column. It should be atleast 10 if it is very less then we need to check the cache.
6. Now go to concurrent:Manager --> Define --> f11 and search for standard manager and see the cache size.
7. If it is very less then change that size to 8 or 9 or 10 and save.8. Also check the workshift --> sleep seconds and processes.
8. keep sleep second to be 30 as average.9. Now check in Concurrent --> Administer -->
9. Check the actual,target and running it should run atleast 10 ata time in the running column of standard manager.
To Administer Concurrent Manager:
To check the Cache Size:
====================================================================================================================
No comments:
Post a Comment