Sysout not updating??

[stevemills]

Hi kids,
I am running CTM6.2 We have a job that started at 2300 and is still yellow as of 0700 yet the sysout has not been updated since 2330.
Is there anything Control-m related that I can look at?? Is this a Control-m issue?? Naturally, my client wants to blame Control-m but I do not think that is true. Any ideas???

[hipikll]

Does your problem touches only this one job or more jobs on this server?
Can you test it with test job. Just to return a output from set command for example.
If it runs ok, then you have to focus on failed job.
If it remain yellow for a long time, you should first check, if the hostnames are known for server and agent (and vice versa).

[th_alejandro]

Usually, this kind of problem is a wrong definition in your program execution flow (not control-m). For example, make sure that the script or program that are executing, never request user keyboard entries. Sometimes, login, or a simply ENTER and the end of the process causes that the job remain yellow, wait for keyboard input.

If your script make ftp via shell (not usinf AFT services from BMC), sometimes when ftp fails and the operating system are configured to automatic retries, the job will remian yellow (running) and not activiti reflect in the sysout, simply Control-M is waiting for a responde from the operating System.

A good practice to detect this problem is define 'Late Time' in the POSTPROC of the job to send message to the ECS console. This can be defined when you know aproximately the average time of execution for the job.

[Walty]

I agree with the previous answers but you can try to investigate about this.
Some suggestions that come to mind:

- How do you see the job through the utility <ctmpsm> : still active?
- If you are connected on the agent, you see this job in executing mode ?
- The definition of Agent to Server port and Server to Agent is correct?
- Do you use agent trough a firewall?
- The tracker agent <p_ctmat> is active?
- Execute <Force Download> to synchronise AJF of Control-M/S and Control-M/EM
- Error messages in directory <proclog> for server and agent?
- The query SQL like 'select JOBNAME, MEMNAME, STATUS, STATE from CMR_AJF where JOBNAME =' your job ' gives you what?

Valid codes for STATUS are:

STATUS = Y means that the job has completed successfully.
Status = N means that the job has not completed or it has completed unsuccessfully.

Valid codes for STATE are:

0 Wait for Scheduling Criteria to be evaluated
1 Wait for confirmation from user
2 Submitted to Agent
3 Not submitted to Agent
4 Executing
5 Ended
6 Analyzed Post Processing Steps
7 Disappeared
8 Post Processed
9 Not Found
A Waiting to Rerun
Waiting for B From Time
C Wait for In Condition
D Wait for Resource
E Wait for Submission
F Not used by CONTROL-M/Server
G Retry to Agent Submission
H Wait for Group Scheduling
I Failed To Submit
J Job waiting for Odate
K Job is past its Odate
Z Agent Status is Unknown

[rahulsehgal]

Hi Steve,

you can try this too,

This may happen when the Control-M agent get unavailable for a minute or two and again comes alive, check the log of the job if you see anything like this:-

ORDERED JOB:1480713; DAILY XXXXX, ODATE 20090723
JOB STATE CHANGED TO Wait Time Window
JOB STATE CHANGED TO Wait Resource
SUBMITTED TO spriXXXXXXXX
STARTED AT 20090724020728 ON spriXXXXXXXX
JOB STATE CHANGED TO Executing
JOB STATE CHANGED TO Unknown
JOB STATE CHANGED TO Executing

If yes then this is because tracker was distracted and was unable to communicate back again to agent properly due to which job is no more executing on the server and no Sysout is generated for the same.

regards,