From time to time we are getting messages similar to the below:
CIAS-04 Job JOBA (0571) errors returned on SQGET Update 0001
or:
CIAS-12 Job JOBA (0571) errors returned on SQGET Update 0001
The jobs themselves finish ok, but what does CA recommend as Best Practice in addition to the explanation reported in the CA7 Message Guide?
This is the description of the message from the CA7 Message Reference Guide:
CIAS-04
Job aaaaaaaa(nnnn) errors returned on SQGET Update nnnnn
Reason:
When trying to retrieve job data (SQGET) for job aaaaaaaa with CA 7 job number nnnn, a nonzero return code nnnn was received.
Action:
Perform the following actions:
? Verify that CA WA CA 7 Edition is responding correctly to other queue requests (such as LQ,JOB=).
? Contact your installation specialist if you have CA WA CA 7 Edition problems.
As long as the involved jobs are completing as expected, this error can be ignored. They occur when the agent sends feedback for a job that has already gone through CA 7 completion processing (e.g. it either completed successfully or has failed and is back in the REQ queue). Since CA 7 does not need the information, it is discarded and an informational message is written to the browse log.
A variation of this message is:
'CIAS-12 Job JOBNAME(nnnn) errors returned on SQGET 0001'.
This commonly occurs when you cancel a job from CA 7 that is already at the agent. When the agent sends feedback, the job cannot be found. If CA 7 is not processing the job correctly, then the message should be investigated and a Case should be opened to Ca Support. In that case the CA 7 log for the time period in question and History 08 report would be needed as first documentation in order the debug the problem deeper.
If you wish to pursue this further (e.g. see the exact message we are discarding) you have two options.
Option 1
1) Determine the time you received the CIAS-04 message
2) Browse the agent transmitter log for that time frame (you may have to look at an archive log). See if there are messages for that time frame. See further below for an example of what a job feedback message looks like. There are different messages returned by the agent.
3) Assuming you find messages. Send to CA Support the agent transmitter log and CA 7 log for the date time the problem occurred with the job name and number to look for to check it further.
Option 2
1) Capture the IAS log at the time the error occurs using the /IAS,FUNC=LOGPRINT,LOGDD=ddname command. Note this log wraps. Depending on your activity, the information needed may or may not be there. If it is not you will need to recycle CA 7 after expanding the log size (see example below)
2) Issue a FIND command for 'Recv Message From:' and look for the job name and number. In the sample below, RHEL01-01 is the agent, RHEL0001 is the job name. and N00009 is the job number.
Recv Message From: RHEL01-01, MsgChar=02, MsgEnc=03, MsgQType=00, MsgQID=7413, MsgNo=00002544, MsgFlag=80,
Msg=RHEL01-01 20160212 21435117-0530 CA75CA75 RHEL01-01 RHEL0001.N00009/RHEL.S00001D160431112/MAIN State FAILED Cmpc(9) SetEnd
Status(Aborted, Signal 9) User(RHEL01-01) Host(WA
3) If you find something similar to the above after the CIAS-04 message time that will be the message IAS is ignoring.
Sample message in the agent transmitter file:
Manager is CC75CC75, Agent is USER1, job name is AGNTJOB1, job number is N00011
Message has been sent: 20160212 10580504+0600 CC75CC75 USER1 AGNTJOB1.N00011/TESTNTWK S00001D160431157/MAIN State COMPLETE Cmpc(0) SetEnd JobLogId(4C665BF7390D08FF7FA627D3235A4305C7D3FC4A) User(USER1) Host(HOSTA)
To increase the size of the internal log add the LOGGER statement. A sample of which is shown below. The default if there is no LOGGER statement is 5MB.
MANAGER +
NAME(CA75CA75) ADDAGENT(100)
AGENTRCV CA75 PORT(47010)
LOGGER SIZE(500) FJOB,JOB=*,FROM=(101023,0600),SPAN=08