Easy question

davidtuti · Post by **davidtuti** » 08 Oct 2009 2:59

Hi,

If there is machine where it's running a job and suddenly the conexion of the network card is down, what Control-M Server and GUI do when the network is up again?

Stop the chain in these job, and the job ended not ok? the chain continue and when Server reconecting with the agent the chain continues?

Many thanks and sorry for my english!

rahulsehgal · Post by **rahulsehgal** » 08 Oct 2009 3:49

Hi David,

Whenever agent(Machine) connection get down job status went to Unknown Status and Control-M Server sends traps to Control-M Agent for Some time like 180 seconds, if agent still not respond on specified time Job status changes to Ended Not Ok state and if agent response back in specified time 180 Seconds, job status changes to executing again.

Regards,
Rahul Sehgal

davidtuti · Post by **davidtuti** » 08 Oct 2009 4:16

Hi!
Many thanks!

But one think, The script in the agent carries on executing until the script finished no?

Thanks!

rahulsehgal · Post by **rahulsehgal** » 08 Oct 2009 6:29

Yes David, it will execute till end.

davidtuti · Post by **davidtuti** » 09 Oct 2009 11:42

Thanks!

Therefore the job it will execute till end and the job shows in red, because the agent not respond in 180 seconds no?

Thanks

rahulsehgal · Post by **rahulsehgal** » 09 Oct 2009 1:29

yes David, you got it, the job will be in failed status on Control-M but the script which is started by job on server will run till end until there is some issue occur in Script coding.

regards,
Rahul Sehgal

hipikll · Post by **hipikll** » 10 Oct 2009 6:26

rahulsehgal
are you sure?

I do not know, if my memory is ok in this area, but:

1) If agent is unavailable - connection is lost, the job is marked as UNKNOWN till the agent is up again or NDP.
2) The script finish to run on agent host. The ctm agent will store the status of the ended 'job' in file and if the connection with server is restored, the job status is read from this file. So the status of the job is not lost.

rahulsehgal · Post by **rahulsehgal** » 12 Oct 2009 11:48

Hi hipikll,

1) If agent is unavailable - connection is lost, the job is marked as UNKNOWN till the agent is up again or NDP.

Same I told before in my previous statement:- "Whenever agent(Machine) connection get down job status went to Unknown Status and Control-M Server sends traps to Control-M Agent for Some time like 180 seconds"

2) The script finish to run on agent host. The ctm agent will store the status of the ended 'job' in file and if the connection with server is restored, the job status is read from this file. So the status of the job is not lost.

You are right in this point but if Server is unable to communicate back to Control-M Agent, job status changes to Ended Not Ok(Failed) from Unknown, If server is able to communicate back then Job status changes back to Executing, till the time server is trying to communicate back to agent job status remains as "Unknown".
this is only the Control-M Part, and about the script Control-M job started on the Control-M agent, this will continue till end no matter agent is down or is trying to communicate with Control-M Server.

Regards,
Rahul Sehgal

davidtuti · Post by **davidtuti** » 29 Oct 2009 6:26

Hi,
how do you know that the time default is 180 seconds? Where could I configure this option? I think that is in ag_menu (server to agent timeout), but in this option I have 120 not 180.

Many thanks!

rahulsehgal · Post by **rahulsehgal** » 02 Nov 2009 2:19

Hi David,

You can update this option under Parameter Customisation of ctm_menu.
I just took 180 second as an example, default for this is 120 seconds.

Regards,
Rahul Sehgal