SyncMirror is a synchronous data mirroring technology between two storage systems or storage components, in which changes are written simultaneously to the primary and mirrored storage. This mechanism is used to ensure high data availability and continuity of critical IT systems in the event of hardware or site failure.
Depending on the vendor and specific solution, SyncMirror may refer to a particular storage technology or feature. The term SyncMirror is best known in NetApp systems, where it is used for synchronous aggregate mirroring and data protection at the disk pool level.
How SyncMirror Works
When SyncMirror is used, two identical copies of the data, called plexes, are created. They are located on different sets of disks or in separate storage pools. All write operations are performed synchronously: the system confirms a successful write only after the required data has been written to both copies.
In simplified terms, the process works as follows:
- The application sends a data write request.
- The storage system sends the changes to the primary and mirrored plexes.
- The data is written to both sets of disks.
- Once the operation is complete, the storage system confirms the write to the application.
This ensures that both copies remain up to date. If one storage component becomes unavailable, the system can continue operating using the remaining copy.
What SyncMirror Is Used For
The primary purpose of SyncMirror is to reduce the risk of data unavailability or loss caused by storage system component failures. The technology is particularly useful for infrastructure where prolonged downtime is critical to business processes.
SyncMirror makes it possible to:
- maintain two synchronized copies of data
- preserve access to data if one set of disks fails
- increase storage system fault tolerance
- reduce the risk of losing the latest changes
- physically separate copies for additional infrastructure protection
Depending on the hardware architecture, the copies may be located in different disk shelves or other isolated storage components.
How SyncMirror Differs from RAID
SyncMirror and RAID address related but different tasks.
- RAID distributes data and redundancy across disks within a RAID group, allowing the system to withstand the failure of a certain number of drives.
- SyncMirror creates a second synchronous copy of the data on a separate set of disks.
RAID primarily provides fault tolerance at the drive level, while mirroring provides an additional layer of protection by separating copies across independent hardware groups.
These mechanisms can be used together: each SyncMirror plex can itself consist of RAID groups. Therefore, SyncMirror does not replace RAID but complements it by providing an additional layer of redundancy.
SyncMirror and Backup
SyncMirror should not be considered a replacement for backup. Synchronous mirroring primarily protects against hardware failures: if one copy becomes unavailable, the other still contains up-to-date data.
However, logical errors can also be synchronously replicated to both copies. For example, file deletion, data corruption caused by an application, or an unwanted change may also be reproduced on the mirror.
Therefore, for comprehensive data protection, SyncMirror is used together with snapshots, backups, and other recovery mechanisms.
Advantages and Limitations of SyncMirror
The main advantages of the technology include high data availability, minimal divergence between the primary and mirrored copies, and the ability to continue operating in certain storage failure scenarios.
At the same time, mirroring requires additional resources: the second copy requires a comparable amount of disk capacity. In addition, SyncMirror does not protect against every data-loss scenario and should be part of a comprehensive fault-tolerance and backup architecture.