13 Replies Latest reply on Oct 20, 2015 5:29 PM by Beauty_and_dBest

    Lots of Concurrent Logs

    Beauty_and_dBest

      Hi ALL,

       

      11i, 12.1, 12.2

      RHEL

       

      Where do I find/location of the paths of all logs in EBS? and which aps table names can I select them?

      We have lots of concurrent managers logs and reports accumulated(in 2 years) in the server.

      Is it okay remove all except a month and less old using rm -f * ?

       

      If I remove it using rm -f *, can I still see them in the form view listed below?

      Capture.PNG

      Or do I need to run purge concurrent program scheduled periodically? (every sunday).

      In your experience with your users, how long in the retention period of the report outputs?

      Can you give me the program name to purge all types of EBS logs?

       

      Please help...

       

      Thanks a lot.

      jc

        • 1. Re: Lots of Concurrent Logs
          Chiranjeevi M A

          Hello JC,

           

          Refer below MOS:

          EBS R12.0.x / R12.1.x : How To Find Location Of Install, Autoconfig, Patching , Clone And Other Logs ? (Doc ID 804603.1)

          Oracle Applications E-Business Suite 12.2 Fusion Middleware Log Files: Locate,View, and Control (Doc ID 1366187.1)

          Managing Concurrent Manager Log and Out Directories (Doc ID 1616827.1)

           

          -Chiranjeevi M A

          1 person found this helpful
          • 2. Re: Lots of Concurrent Logs
            Beauty_and_dBest

            Thanks

             

            Checking the docs now.

            • 3. Re: Lots of Concurrent Logs
              Bashar.

              Hi,

               

              You should purge your concurrent requests periodically (daily is preferable).

               

              Run the "Purge Concurrent Request and/or Manager Data Program" request with the Mode "All" and Age "30" parameters to clean all requests submitted more than 30 days ago.

              Some companies purge requests older than 7 days but I think that this leaves little history to diagnose some issues.

               

              No need to purge files from the file system unless you are running a very old release (11.5.5 or below)!

               

              Regards,

              Bashar

              1 person found this helpful
              • 4. Re: Lots of Concurrent Logs
                Beauty_and_dBest

                Thanks Bashar and ALL,

                 

                ****"Purge Concurrent Request and/or Manager Data Program" request with the Mode "All" and Age "30"


                Will the above program includes the SYSADMIN submitted request like "Gather Schema Stats",  "OAM dashboard", etc? as shown in the above request queue?

                Or it is for user reports only?


                We have 50Gb worth of reports output to be purged. Will it not bug down the instance to purge all of these reports in one time run

                 

                Thanks.

                • 5. Re: Lots of Concurrent Logs
                  Pravin Takpire

                  First start with Age of 365 -- it will delete all old logs before 1 year

                  and then keep on reducing to acceptable days to keep logs e.g 30

                  and then schedule the same request to run everyday in night where load is less

                  regards

                  Pravin

                  1 person found this helpful
                  • 6. Re: Lots of Concurrent Logs
                    Beauty_and_dBest

                    Thanks Pravin and ALL,

                     

                    Are request output logs submitted by SYSADMIN like "Gather Schema Stats",  "OAM apps dashboard" included in the purge program?

                     

                    Kindest regards.

                    • 7. Re: Lots of Concurrent Logs
                      Pravin Takpire

                      yes

                      1 person found this helpful
                      • 9. Re: Lots of Concurrent Logs
                        Beauty_and_dBest

                        Hi ALL,

                         

                        I run  the purging of logs with 40Gb total size, but I got error bellow:

                        Capture2.PNG

                        +---------------------------------------------------------------------------+

                        Application Object Library: Version : 12.0.0

                         

                        Copyright (c) 1979, 1999, Oracle Corporation. All rights reserved.

                        FNDCPPUR module: Purge Concurrent Request and/or Manager Data

                        +---------------------------------------------------------------------------+

                        Current system time is 20-OCT-2015 21:01:50

                        +---------------------------------------------------------------------------+

                         

                        APP-FND-01564: ORACLE error 1555 in request

                        Cause: request failed due to ORA-01555: snapshot too old: rollback segment number 18 with name "_SYSSMU18_1232623801$" too small.

                        The SQL statement being executed at the time of the error was: and was executed from the file .

                        +---------------------------------------------------------------------------+

                        Start of log messages from FND_FILE

                        +---------------------------------------------------------------------------+

                        +---------------------------------------------------------------------------+

                        End of log messages from FND_FILE

                        +---------------------------------------------------------------------------+

                         

                        +---------------------------------------------------------------------------+

                        Executing request completion options...

                         

                        Output file size:

                        200550889

                         

                        Finished executing request completion options.

                        +---------------------------------------------------------------------------+

                        Concurrent request completed

                        Current system time is 20-OCT-2015 21:35:13

                        +---------------------------------------------------------------------------+

                         

                        Do I need to restart the database to release the undo tablespace?

                        Or do I just need to submit again the program?

                         

                        Thanks a lot,

                        jc

                        • 10. Re: Lots of Concurrent Logs
                          Pravin Takpire

                          you may need to change the age parameter. start with deleting of just 2 months data. By giving age as 700 days, and then gradually go on deleting the data.

                          • 11. Re: Lots of Concurrent Logs
                            Beauty_and_dBest

                            Thanks, But we have 2 years worth of data

                            Can I query it first in the data dict, how much size of logs each month? and how many months did it accumulate? Something like demography.

                            • 12. Re: Lots of Concurrent Logs
                              Bashar.

                              As Pravin already said, you should split the purging into smaller ranges to avoid the snapshot too old error.

                               

                              Here is a query to get the space used of the requests' files per month:

                               

                              select to_char(fcr.request_date,'RRRR-MM') req_month, round(sum(ceil(ofile_size/4096)/256),1) outfile_total, round(sum(ceil(lfile_size/4096)/256),1) logfile_total

                              from fnd_concurrent_requests fcr

                              group by to_char(fcr.request_date,'RRRR-MM')

                              order by 1;

                               

                              Regards,

                              Bashar

                              1 person found this helpful