in HarvestController Server. The response is sent to a queue and is then picked up by a random harvester, which returns it to the queue if the it was meant for a different harvester. Why not use a unique reply queue, or even a Topic so that every harvester reads the reply exactly once. Also the call "close()" doesn't shut the HaCo down.
Checklist
Activity
Show:
Colin Samuel RosenthalOctober 29, 2018 at 1:44 PM
It seems that this bug, or something similar, also arises if you restart the HarvestJobManager. See for example https://kb-dk.atlassian.net/browse/NARK-1426 . The HJM was restarted but harvest could not be scheduled because of
See the method
in HarvestController Server. The response is sent to a queue and is then picked up by a random harvester, which returns it to the queue if the it was meant for a different harvester. Why not use a unique reply queue, or even a Topic so that every harvester reads the reply exactly once. Also the call "close()" doesn't shut the HaCo down.