We are currently upgrading from a standalone Exchange 2003 server to an Exchange 2007 CCR environment. There is one server with the CAS/HUB role and two servers playing the active and passive nodes of the clustered mailbox server role.
At the moment, we only have enough free space to host mailbox and public folder databases on one of the nodes. The database paths on the passive node will be set up after the migration has been completed and the free space recovered and reused from the Exchange 2003 server.
The issue with this scenario is that a large chunk of transaction logs are building up on the active node because it cannot replicate the transactions to the passive node. Even though we are performing nightly backups, the logs are not being removed because of the fact that they have not been committed to the database on the passive node (which obviously doesn’t exist).
To be on the safe side, the logs are backed up in case they have to be replayed on the passive node after the storage has been provisioned. However, I am hoping there is a better solution. Once the migration has been completed, we will have approximately two ~200GB databases on the active node. Rather than restoring the logs from backup and replaying them on the passive node, is it possible to take the databases down, copy the physical .edb files from the active node to the passive node, then bring them back up? Can I be 100% sure that the copy status of the databases after this operation will be fully synchronized?