After an upgrade to Performance Management the Data Aggregator intermittently fails to synchronize successfully with Performance Management.
In the problem environment there is also a 'hot standby' Performance Management environment where the Performance Management system is running at the same time as the active production environment.
All supported Performance Management releases
Production Data Aggregator was communicating with the Disaster Recovery system.
A 'hot standby' Disaster Recovery system isn't supported. It is known to cause many problems, cross environment communication being at the root of many of them.
Ensure the different environments are speaking with each other only and not communicating between each other.
While unsupported, if this 'hot standby' environment is a requirement, one possible way to avoid issues like this is the use of iptables and/or firewall rules. Enable them during regular usage to prevent the systems from communication with each other. When communication between the environments is needed, like to fail over, to transfer DB backups or running Data Repository copycluster cycles, disable the rules and re-enable them when done.
Performance Management Disaster Recovery documentation