OverviewExinda in a cluster will not necessarily have the same size database. This means the monitor partitions will not be the same size.
Each device has its own database, and they are sending each other flows via the cluster link.
Some of the reasons why the database sizes are different.
1. When the devices get into a cluster they do not back fill each other's databases with their historical data. The syncing is from when the clustering begins. So this results in different sizes of databases.
2. Different table sizes on the databases. This is because the different devices in a cluster had different database before the clustering.
3. If the cluster link went down at some point then during this time the data being inserted into the databases will be different.
4. If one of the devices was very busy and couldn't accept the other devices flows, so those flows were dropped then this increases the differences in sizes. A good example of this is during a SYN flood one of the devices is so busy that it cannot process flows from its peer.
Unless there is database errors in the logs, the best option is to leave the databases the way they are.