6 Replies Latest reply: Jun 20, 2014 4:24 PM by user8860348 RSS

    Process Scheduler Boot Error

    user8860348

      Folks,

       

      Hello. I am running Oracle Database 11gR1 with Operating System Oracle Linux 5.

      My PeopleSoft Internet Architecture is WebLogic/Tuxedo/PeopleTools8.53.

       

      After I boot Application Server with 8 processes sucessfully, I boot Process Scheduler and get the error below:


      Booting admin processes ...

      exec BBL -A :
              CMDTUX_CAT:822: ERROR: No BBL available, cannot boot

      tmboot: WARN: No BBL available on site linux.domain.
              Will not attempt to boot server processes on that site.


      tmboot: CMDTUX_CAT:827: ERROR: Fatal error encountered; initiating user error handler

      exec tmshutdown -qy


      tmshutdown: internal error: CMDTUX_CAT:764: ERROR: can't attach to BB


      tmboot: CMDTUX_CAT:828: ERROR: Cannot create error process - fork() failed

       


      I have checked the current server log file SCHDLR_0618.LOG and no message in it.

      I have checked the current Tuxedo log file TUXLOG and its content is below:


      Begin attempt on domain HRCS90
      105113.linux.domain!tmadmin.8296.2885232272.-2: TMADMIN_CAT:1330: INFO: Command: boot -A
      105118.linux.domain!BBL.8298.2841136176.0: 06-18-2014: Tuxedo Version 11.1.1.3.0, 64-bit, Patch Level (none)
      105118.linux.domain!BBL.8298.2841136176.0: LIBTUX_CAT:681: ERROR: Failure to create message queue
      105118.linux.domain!BBL.8298.2841136176.0: LIBTUX_CAT:248: ERROR: System init function failed, Uunixerr = : msgget: No such file or directory
      105118.linux.domain!BBL.8298.2841136176.0: CMDTUX_CAT:26: INFO: The BBL is exiting system
      105118.linux.domain!tmboot.8297.3527355488.-2: 06-18-2014: Tuxedo Version 11.1.1.3.0, 64-bit
      105118.linux.domain!tmboot.8297.3527355488.-2: CMDTUX_CAT:825: ERROR: Process BBL at linux.domain failed with /T tperrno (TPEOS - operating system error)
      105118.linux.domain!tmboot.8297.3527355488.-2: WARN: No BBL available on site linux.domain.
              Will not attempt to boot server processes on that site.
      105126.linux.domain!BBL.8304.4191628336.0: 06-18-2014: Tuxedo Version 11.1.1.3.0, 64-bit, Patch Level (none)
      105126.linux.domain!BBL.8304.4191628336.0: LIBTUX_CAT:334: ERROR: No BBL
      105126.linux.domain!BBL.8304.4191628336.0: LIBTUX_CAT:248: ERROR: System init function failed, Uunixerr =
      105126.linux.domain!tmboot.8303.4140594864.-2: 06-18-2014: Tuxedo Version 11.1.1.3.0, 64-bit
      105126.linux.domain!tmboot.8303.4140594864.-2: WARN: No BBL available on site linux.domain.
              Will not attempt to boot server processes on that site.
      105126.linux.domain!tmboot.8303.4140594864.-2: tmboot: CMDTUX_CAT:827: ERROR: Fatal error encountered; initiating user error handler
      105126.linux.domain!tmboot.8303.4140594864.-2: FATAL: CMDTUX_CAT:828: ERROR: Cannot create error process - fork() failed

      105126.linux.domain!PSADMIN.8274: End boot attempt on domain HRCS90


      My question is:

      As you see above, CMDTUX_CAT:825 could be Operating System error or the Application Server has outgoing.

      Can any folk help to solve Process Scheduler boot error ?


      Thanks in advance.

        • 1. Re: Process Scheduler Boot Error
          BobLilly

          Most likely your issue is related to the following messages:

           

          105118.linux.domain!BBL.8298.2841136176.0: LIBTUX_CAT:681: ERROR: Failure to create message queue

          105118.linux.domain!BBL.8298.2841136176.0: LIBTUX_CAT:248: ERROR: System init function failed, Uunixerr = : msgget: No such file or directory

           

          The second message is most likely a result of the first message. It could be a permissions issue (although this is less likely if the app server starts ok). Review the instructions in Chapter 10B of the installation guide and make sure all of your environment settings are correct. Verify that you are on a certified OS/DB platform (must be 64-bit version of Oracle Linux 5). Check note 1575892.1 for patch requirements (minimum 5.6 for Oracle Linux 5). Check support note 608218.1 and make sure your OS kernel settings are ok.

           

          Check your Process Scheduler configuration for incorrect paths or, if necessary, increase the log fence setting to 5, try starting it again, then check the tuxlog to see if the increased logging provides any helpful information.

           

          There is a known issue on Windows platforms when trying to start the process schedulers from a remote desktop session, but I'm not sure how it might apply to a linux environment. The error message is the same, though.

           

          Regards,

          Bob

          • 2. Re: Process Scheduler Boot Error
            user8860348

            Folks,

             

            Hello. Thanks a lot for replying. I have checked Linux parameter as below:

             

            # more /proc/sys/kernel/msgmni

            It returns 16.

             

            I increase its value form 16 to 1024 by adding the line below into the file /etc/sysctl.conf :

            kernel.msgmni = 1024;

             

            After that, I reboot Linux and Process Scheduler. In TUXLOG file, the 4 errors below are gone:

            LIBTUX_CAT: 681

            LIBTUX_CAT: 248

            CMDTUX_CAT:825

            LIBTUX_CAT:334

             

            While the above 4 errors are gone, a new error comes up as below:

            LIBTUX_CAT:250:ERROR: tpsvrinit( ) failed.

            tmboot: CMDTUX_CAT:827:ERROR: Fatal error encountered; Initiating user error handler.

             

            The above new error means Application Server cannot locate my Database. I have checked factor by factor :

            1) Tuxedo environment parameters: a few parameters are the same when Process Scheduler boots 7 processes successfully.

            2) I can connect with my Database Sever from another machine successfully.

            3) My Database path is correct.

             

            I cannot find out why Database cannot be located by Application Server and causes tpsvrinit( ) failed.

             

            My question is:

            Can any folk help to solve the error LIBTUX_CAT:250:ERROR:tpsvrinit( ) failed ?

             

            Thanks in advance.

             


            • 3. Re: Process Scheduler Boot Error
              2630053

              1. Please check and confirm that you have configured the kernel for enough shared memory segments and sizes.

              2. If you are suspecting not connecting to DB, then enable tracing and see where it failing.

              3. Ensure your starting account and environment variables are properly configured.

              4. Assuming you are starting PRCS on the same machine, I am seeing a bit of conflict - u had started thread with process scheduler but in your second reply, Application Server is referenced.

              5. full log will help to isolate the problem instead of portions of it.

              • 4. Re: Process Scheduler Boot Error
                user8860348


                Folks,

                 

                Hello. I have configured the Report Node in Browser. Process Scheduler boots 8 processes successfully now.

                 

                But when I run reports using Process Scheduler in Browser, all of the reports are success but not posted and get the same error message as below :

                 

                Error converting HTTP response to UCS2.

                XML document object creation failed.

                Unable to process HTTP reply from Report Repository.

                HTTP transfer error.

                 

                According to Nicolas' tutorial http://gasparotto.blogspot.com/2010/04/troubleshooting-process-scheduler-and.html regarding th above error, "The local node must be trusted, and password authentication must be set." I cannot find the Navigator in browser to do this.

                 

                My question is:

                How to solve the above "HTTP transfer" error ? What is the Navigator path to set the local node and password authentication to solve the error ?

                 

                 

                Thanks in advance.

                • 5. Re: Process Scheduler Boot Error
                  user8860348

                  Folks,

                   

                  Hello. The navigator is PeopleTools>Integration Broker > Integration Broker setup > Nodes

                   

                  I have changed authentication option from none to password. I reboot Process Scheduler and get a new error as below:

                   

                  Client error Http Reply Code: 401 - Unauthorized. HTTP Transfer Error.

                   

                  This error means UserID and Password are not right.

                  Let me explain a few password as below:

                   

                  My Integretion Broker Node password: myname

                  My Report Node Login ID and Password: PS and myname

                  My Browser PeopleSoft Application Signon User ID and Password: PS and myname

                  My OS Linux login ID and PAssword: myUserId and myname

                   

                  The above 4 system logins have the same password: myname

                   

                  I don't know what causes "Client error Http Reply Code:401 - Unauthorized."

                   

                  My question is:

                  Can any folk help to solve the issue: Process Scheduler runs report error: "Client error Http Reply Code: 401-Unauthorized. HTTP transfer error" ?

                   

                  Thanks in advance.

                  • 6. Re: Process Scheduler Boot Error
                    user8860348

                    Folks,

                     

                    Hello. No matter what happened, Process Scheduler has booted 8 processes successfully. Let's end this thread and start with a new discussion.