Local Continuous Replication is an Exchange 2007 feature that provides data redundancy on a single server. The server keeps two copies of the databases so that if the primary set of databases suffer a failure due to virus scanning or hardware failure the secondary copy can be moved into place. You might think that you could just put a second disk in your server and but the passive copy of the database on it, however this misses the full purpose of LCR. Since database loss often is the result of disk, controller or other hardware failing, its best to isolate the active and passive databases as much as possible from each other. It is recommended that the databases are not only on separate disks but that those disk are attached to different storage controllers. A simple example of a LCR configuration is in the picture.
Configuration considerations
Since both the active and passive database must be maintained by a single server, there will be additional CPU (~ 20% more) and memory (1GB or more) overhead when you are using LCR. Also, as will the CCR and SCR only one database can be created in each storage group that is LCR enabled.
When would you use LCR? It is an excellent solution for smaller offices that do not have clustering experience but are concerned data redundancy.
Pros:
- Inexpensive data redundancy
- Server running LCR can also host other Exchange roles (hub transport, unified messaging, Client Access)
- Increase recommended maximum database size from 100 GB to 200GB
- Improved recovery time for database failures
- Does not require Microsoft clustering or Windows Server Enterprise Edition.
Cons:
- Increased memory and processor load on server
- Failover is not automatic to passive database
- Cannot use LCR on a public folder database if there are other public folder servers in the Exchange organization.
You can find information about installing LCR from my previous post, Installing an Exchange LCR Sever step-by-step.
Posted
Nov 09 2007, 03:58 PM
by
Joel Stidley