Jobs in Time Overrun with Uproc nº: 0000000

book

Article ID: 130250

calendar_today

Updated On:

Products

CA Automic Dollar Universe

Issue/Introduction

A week before the DST time in Europe ( since 24th March 2019), some Jobs scheduled by a Task go straight to "Time overrun" status when they are supposed to be Launched (Start date)  instead of waiting until the end of the Launch Window (End date).
They can be easily identified as they have as Uproc nº 0000000

When trying to display the History Error, a popup error appears in Univiewer Console as the History Trace was not created:

<Please see attached file for image>

Error in UVC - History Trace

There is no reason for this issue as the Launcher was up and running, and the universe.log does not contain any Error.

 

Cause

Cause not identified for the moment but seems linked to the DST Summer Time that occured the 31st March 2019 in Europe.

Environment

The issue has only been observed on AIX systems running Dollar Universe 6.7.41, 6.9.01 and 6.9.41 in European Servers.

Resolution

In case the issue occurs before another DST, the workaround consists on Updating the Jobs in Time overrun, two options are possible:

a) Manual Action for Operators via UVC.
In Job Runs, click right on the affected Job and click on Update.
Modify the "End date" so that if contains a date in the future, and click on Run.

<Please see attached file for image>

rerun time overrun jobs

b) Automatic action via u_spv_com.ksh script to be triggered via an Alerting Rule.

1. Create and Enable via UVC - Operations Settings the following Alerting Rule or modify the existing one so that it is launched for all jobs in Status Time Overrun.

<Please see attached file for image>

alert_rule

2. Then edit the script u_spv_com.ksh of the impacted node/s via Node Settings - Node Files:

<Please see attached file for image>

modification of u_spv_com.ksh

Add from line 64 the following:
# ============================================================================================================= # 
# CA a Broadcom company # 
# PROCEDURE: Broadcom's Workaround for Incident 01329930 # 
# SUBJECT: Captures Jobs incorrectly ended in Time Overrun # 
# And restarts automatically the Job # 
# Checks if Job Status=TIME_OVERRUN and Uproc number=0000000# 
# ============================================================================================================= # 
if [ $DU_STATUS="TIME_OVERRUN" ] && [ $DU_NUMPROC=0000000 ] 
then 

$UNI_DIR_EXEC/uxupd fla upr=$DU_UPROC mu=$DU_MU nseq=$DU_NUMLANC now 

fi 
#===================== END ================================== 


Example:

<Please see attached file for image>

fix in u_spv_com.ksh

3. Save and close, the script u_spv_com.ksh should have execution rights for all users.

-rwxr-xr-x. 1 univa univa 3902 Mar 29 11:07 u_spv_com.ksh

4. In case of new occurrence, the Job should be automatically relaunched.


IMPORTANT:
Issue ( Jobs in Time Overrun) did not occur anymore after the Summer DST (31st March 2019).
The Alerting Rule should not remain Enabled after that.



  

Additional Information

The issue was indeed related to the DST time change and did not occur anymore after the time change occured on 31st March 2019 at 3am

Attachments

1558687957552000130250_sktwi1f5rjvs16fg4.png get_app
1558687955696000130250_sktwi1f5rjvs16fg3.png get_app
1558687953930000130250_sktwi1f5rjvs16fg2.png get_app
1558687952235000130250_sktwi1f5rjvs16fg1.png get_app
1558687949051000130250_sktwi1f5rjvs16fg0.png get_app