Details
-
Fix
-
Status: Released (View Workflow)
-
Minor
-
Resolution: Fixed
-
2.7.0
-
None
Description
Current Situation
If the Subagent is stopped during execution and started again, it becomes decoupled. When the Subagent is reset, the following error message is logged in the Primary Director agent log file:
2024-09-12T11:39:34,803 ERROR js7.subagent.director.SubagentKeeper - Subagent:subagent-unix reset => UnknownEventId: An unknown EventId has been requested: 1726133744783000/2024-09-12T09:35:44.783Z, oldest EventId is after 0/BeforeFirst, newest is 1726133768130000/2024-09-12T09:36:08.130Z
Desired Behavior
If the Subagent stops during execution, when it starts again, it should attempt to couple with the Primary and Secondary Director agents automatically, without requiring a reset or forced reset process.
Workaround
If the Subagent stops during execution, follow these steps:
- Check which agent is decoupled and verify if any process is running on that agent.
- If no process is running, reset force the Subagent from the Manage Controllers/Agents section and wait some time. Afterward, the agent should couple again.
How to Reproduce
Follow these steps:
- Add an ad hoc order in the Workflow that runs on the Subagent.
- Stop the Subagent service, wait for 30 seconds, and start the Subagent service again.
- Use the following command: bin/agent_<port>.sh stop && sleep 30 && bin/agent_<port>.sh start
- Go to the Resources tab and check if the Subagent status is showing "decoupled."