AggregationTimeoutChecker does not transfer to failover camel instance

Sample camel scenario:

• jms queue (exclusive consumer) -> aggregation (jdbc) -> jms queue
• aggregation onCompletionTimeout = 60 seconds
• Two camel route containers in a master/slave configuration

If a message enters the aggregation jdbc data store and the master camel instance fails, the second camel instance will ignore existing message in the aggregation store but continue to process new messages from the jms queue correctly.

When the failover occurs, the second instance should pick up the existing aggregation messages, similar to what happens when the master is re-started (e.g. "Restored 1 CompletionTimeout conditions in the AggregationTimeoutChecker").