Hi,
Has anyone seen this before?
We have jobs that have been submitted to a Control-M server that remain in an executing state. I have tested the agent comms and all is in order. I can succesfully ping the server and the agent is listed as available. When jobs are uploaded to the Control-M Server they are listed in a wait state and when I hold the jobs, release the job, it remains in a executing state and that state never changes. I have changed the job to a normal 'ls', still no luck. The server used to work, now it does not.
Andrew
Jobs remain in Executing state
Hi Andrew.
Yes, I have seen already that issue. It's a bug already referenced by BMC.
In fact they give the following solution.
1- Hold your job
2 - Launch SQL command. (ust ctm_menu)
Find your job in CMR_AJF using that SQL command, for example:
select JOBNO, JOBNAME, MEMNAME, HOLDFLAG, STATUS, STATE from CMR_AJF where HOLDFLAG='Y' and ODATE='20070117'
3 - Update your job with that informations
update CMR_AJF set STATUS='N', STATE='8' where JOBNO='your JOBNO
4 - Free your job. It will appear in Red.
It could be the same with a Blank Job.
Yes, I have seen already that issue. It's a bug already referenced by BMC.
In fact they give the following solution.
1- Hold your job
2 - Launch SQL command. (ust ctm_menu)
Find your job in CMR_AJF using that SQL command, for example:
select JOBNO, JOBNAME, MEMNAME, HOLDFLAG, STATUS, STATE from CMR_AJF where HOLDFLAG='Y' and ODATE='20070117'
3 - Update your job with that informations
update CMR_AJF set STATUS='N', STATE='8' where JOBNO='your JOBNO
4 - Free your job. It will appear in Red.
It could be the same with a Blank Job.
Jobs remain in Executing state
Hi Andrew,
I meet this situation too in a regular way.
The solution of Franck applies perfectly, but I allow myself to add this complement:
1 - After the update SQL, the job appear in red, it is necessary to execute a 'Force-Ok' to release the resources QUANTITATIVE or CONTROL if used.
2 - Execute a 'Force Download' (ctm_menu/opt 10/opt 16) or command line 'reset_ecs' (utility Control-M) to force the syncronisation Control-M/Server & Control-M/EM
- You will note that the command 'kill job' is without effect and that if you check the execution of the job which is in status executing on the server/agent certainly no jobs are active.
Here other little similar situations met:
1 - Job not started whereas all the criteria of execution are filled and the 'why' command indicates simply one message like 'Performing why'
After investigation it is necessary to update manually (SQL) the level of the 'resources' or 'status' of the jobs.
2 - Resources not released at the end of the jobs (post already put on this forum)
I also advise you to look the status of job by the utility 'ctmpsm' to see whether with Control-M/Server if you have the same situation. Sometimes to solve the situation with this utility without SQL update.
In my environment composed of several server in version 6.1.03 running under (Solaris, Oracle) or (W2003, MS-SQL) and one in version 2.24.05/5.08 running under (Solaris, Sybase) I noted that this occurs in a regular way with Solaris/Oracle, from time to time with W2003/MS-SQL and practically never with Solaris/Sybase (old version).
Regards
Walty
I meet this situation too in a regular way.
The solution of Franck applies perfectly, but I allow myself to add this complement:
1 - After the update SQL, the job appear in red, it is necessary to execute a 'Force-Ok' to release the resources QUANTITATIVE or CONTROL if used.
2 - Execute a 'Force Download' (ctm_menu/opt 10/opt 16) or command line 'reset_ecs' (utility Control-M) to force the syncronisation Control-M/Server & Control-M/EM
- You will note that the command 'kill job' is without effect and that if you check the execution of the job which is in status executing on the server/agent certainly no jobs are active.
Here other little similar situations met:
1 - Job not started whereas all the criteria of execution are filled and the 'why' command indicates simply one message like 'Performing why'
After investigation it is necessary to update manually (SQL) the level of the 'resources' or 'status' of the jobs.
2 - Resources not released at the end of the jobs (post already put on this forum)
I also advise you to look the status of job by the utility 'ctmpsm' to see whether with Control-M/Server if you have the same situation. Sometimes to solve the situation with this utility without SQL update.
In my environment composed of several server in version 6.1.03 running under (Solaris, Oracle) or (W2003, MS-SQL) and one in version 2.24.05/5.08 running under (Solaris, Sybase) I noted that this occurs in a regular way with Solaris/Oracle, from time to time with W2003/MS-SQL and practically never with Solaris/Sybase (old version).
Regards
Walty
-
kennedyk
Unfortunately not only, and it depend on your operating system.Hi,
Thats correct. The problem is mostly caused by the Agent directory being 100% full. This must always be the first place to check when jobs remain yellow
You must check to I/O Disk access, If your disk is used intensively it could be an issue when your agent want to access on it.