What is synchronization?
Synchronization is the process by which Data Protection Manager (DPM) transfers data changes from a protected file server to a DPM server, and then applies the changes to the replica of the protected data. DPM relies on synchronization to keep replicas synchronized with the protected data on the file servers.
You can choose between two modes of synchronization: Only before each shadow copy is created or Nearly-continuous (hourly). If you are protecting data that changes frequently throughout the day, you may want to synchronize your data hourly. If you are protecting data that changes less frequently, you can schedule fewer shadow copies and synchronize your data only before each shadow copy is made. For more information, see How to modify protection options.
DPM provides two methods for synchronizing a replica: incremental synchronization and synchronization with consistency check.
Incremental synchronization (also referred to as synchronization) transfers changes to protected data from the file server to the DPM server and then applies the changes to the replica. When you create a protection group, you specify a synchronization schedule or accept the default schedule. In general, you can rely on incremental synchronization to keep a replica consistent with its data sources. This method is faster and more efficient than performing a consistency check because only data changes since the last synchronization are transferred and applied to the replica.
Synchronization with consistency check (also referred to as “a consistency check”) transfers the data changes from the protected file server to the DPM server but also performs block-by-block verification to ensure that all the data on the replica is consistent with the protected data. This process is slower than synchronization because all the data on the replica is compared rather than simply applying the data changes to the replica. A consistency check impacts the performance of both the file server and the DPM server to varying degrees, depending on network load, CPU processing power, and timing. If you schedule a daily consistency check, you should schedule it when other network traffic is low.
DPM automatically initiates a consistency check:
When a replica is created for a new protection group
When you protect new data sources by adding them to an existing protection group
As a daily task, if you have scheduled a daily consistency check for a protection group
Because of the processing load imposed on both the DPM server and the file server, you should only perform a manual consistency check when it is necessary to make a replica consistent with its data sources. This could happen, for example, when the synchronization log runs out of disk space or when a file server is shut down unexpectedly during synchronization.
A synchronization log is a hidden file, stored on a protected volume, in which the DPM file agent records changes to data since the last synchronization job. The synchronization log makes it possible to incrementally synchronize a replica, thereby speeding up the synchronization process and minimizing the impact of synchronization activity on the network. You can increase or decrease the allocated disk space for the synchronization log, as needed.
A transfer log is a file, located on the DPM server, that stores pending changes to a replica. When a file agent transfers the changes in the synchronization log from the file server to the DPM server, the transfer log maintains these changes until they can be applied to the replica. You cannot modify the allocated disk space for the transfer log. DPM automatically adjusts the space allocated for the transfer log, based on the other disk allocation settings.
- What is a replica?
- What is a consistency check?
- Consistency check in progress
- Replica is inconsistent
- What is a shadow copy?
- How to synchronize a replica
- Synchronization failures