JOBF get stuck waiting for agent even if agent is alive again.

Document ID : KB000089966
Last Modified Date : 14/04/2018
Show Technical Document Details
Issue:
JOBF get stuck waiting for agent even if agent is alive again.
Resolution:

Symptoms

Agent disconnects and reconnect again.

JOBF can not continue to run after the agent was restarted 

07 - 20140425/142125.090 - U0000062 Cannot continue after connection interruption. 

When jobs tried to restart: 

10 - 20140425/144015.605 - U0011000 'JOB.TIMECARDPUBLIHER.COPY_FILES_TO_TEMP_FOLDERS_ON_FLEXIM' (RunID: '0041119834') could not be started on 'VELAX_LINUX'. See next message. 
10 - 20140425/144015.607 - U0011021 Host 'VELAX_LINUX' is not active! 
10 - 20140425/144015.621 - U0011113 Task with RunID '0041119834' will be re-started as soon as host 'VELAX_LINUX' is active again or the previous error has been corrected. 

 


 

Cause

This could be due to one option in the variable UC_HOSTCHAR_DEFAULT ( or UC_HOSTCHAR_agentname) - BLOCK_ON_LOST 

BLOCK_ON_LOST 

This entry serves to define the action that should be taken by ProcessFlows if one of their Jobs changes its status to ENDED_LOST. 

Allowed values: "YES" (default value) and "NO" 
"YES" - The ProcessFlow is blocked 
"NO" - The ProcessFlow is not automatically blocked 

The JOBP set this status because of the LOSTCONNECTION to the agent and so the JOBP got this status for the task and can not resume after agent restarted. 

 


Resolution

BLOCK_ON_LOST = NO (No agent restart required) 

After setting the parameter to "NO", if the job was set to status "ENDED_LOST" or undefined because of the agent aborted prematurely or was canceled then its workflow would not be BLOCKED but continue to run.