At the moment, an agent service is considered ready when the respective service task enters AGENT_EXECUTING state. That is a problem for services which need some time for setup. We should allow to hook back into the control channel for the service to send a ready message to - the agent would then wait for that message instead of a state change.
At the moment, an agent service is considered ready when the respective service task enters
AGENT_EXECUTING
state. That is a problem for services which need some time for setup. We should allow to hook back into the control channel for the service to send aready
message to - the agent would then wait for that message instead of a state change.