Customer has a virtual server <server_Name> which is running on an F11 load-balanced link <hostname>.<domain>.com at the backend, where it points to two servers in a round-robin fashion.
When executing the job through Autosys the job events are moving to the load balancing link.
However, the events are getting distributed in round-robin fashion and the jobs processes are getting pilled up on the agent machines.
Currently, we do not officially support using any type of load balancer in front of agents.
Agents behind an external load balancer will have the following problems:
1. Unreliable ability to cancel jobs
2. Unreliable ability to retrieve job logs
3. Chase failures
We do NOT recommend customers place the agents behind a load balancer for these reasons.
I would suggest you set up each note as a real machine in Autosys and then create a Virtual machine and assign the two nodes to the virtual machine and use the new virtual machine definition in your job.
If you would like to see this added to a future release you will need to open an enhancement request on our idea wall so that our product management team can consider it.