If I understand your question properly, the answer is that the agent will always try to find some communication path to become managed. This is to help prevent the “orphaned agent” issue that chronically will affect many agent based technologies.
Specifically, if an agent is using manual selection, the agent will try the following:
Try its primary relay.
Then if failed, try the secondary relay.
Then if failed, try the “Failover relay” (if the failover relay is defined).
Ben, after step 4, if the agent can’t connect directly to the BES Server, will it go to step one ?
What are the delays between each step ?
If a agent uses its failover relay, due to connection problem with relay 1 and 2, will it try again connection through relay 1/2 later ? If yes, when ?
The agent will wait “_BESClient_RelaySelect_MinRetryIntervalSeconds” (default 1 minute) after completing step 4 and then startover at step 1. Each time it completes a check, it will double its time until “_BESClient_RelaySelect_MaxRetryIntervalSeconds” (default 2 hours) is reached, and it will retry at that interval.
If it finds a relay, it will check again after “_BESClient_RelaySelect_IntervalSeconds” (default 6 hours) to see if there is a better relay choice.