Beta Feedback

Modification to Agent Initial Start Connection Backoff Behavior

This request is to modify the way the agents behave when they initially start prior to ever connecting to an EPM server. This is the behavior as I understand it today: Current Behavior: Agent was just installed and hasn’t reported yet, if there is no network connection or EPM is down the agent will try and connect, fail and go into a back off period (60) seconds and try again. Each time the agent tries to connect and fails the back off period gets longer until it reaches its max connection “Backoff has reached it's maximum of 3600000”or 1 hour and then the watchdog shuts down. At this point the connection will not happen again until the machine is restarted or the process is started manually. Requested Behavior: The request is to have the agent not shutdown the watchdog at the end of repeated failures to connect, and instead, go into a state where it will attempt to connect on an ongoing basis--at least once every 12 / 24 hours. Justification / Reason for Request: The reason this is being requested is for POC's. For POC's we have begun providing agents prior to having a server spun-up. There are multiple reasons behind this including reducing the time it takes to execute POC's. When providing agents ahead of time, clients may install these ahead of the server. This would cause them to attempt to connect and then go into a shutdown state. Many client machines are not rebooted on a continual basis. If the agent watchdog shuts down, it would cause the workstations to not connect which would be detrimental to the overall POC. Having the watchdog go into a state where it checks every 12 / 24 hours would be preferred.
  • Gravatar Troy Partain
  • Sep 11 2015
  • Planned