Disk mirroring, also known as RAID 1, is the replication of data to two or more disks. Disk mirroring is a good choice for applications that require high performance and high availability, such as transactional applications, email and operating systems. Disk mirroring also works with solid state drives so “drive monitoring” may be a better term for contemporary storage systems.
Because both drives are operational, data can be read from them simultaneously, which makes read operations quite fast. The RAID array will operate if one drive is operational. Write operations, however, are slower because every write operation is done twice.
Disk (drive) mirroring is particularly advantageous for disaster recovery scenarios as it provides instantaneous failover for data required by mission-critical applications. If primary drives in the array are damaged or unable to operate, traffic is switched to secondary or mirrored backup drives. The mirror copy is able to become operational on failover because the operating system and application software are replicated to the mirror along with the data used by the applications.
RAID and RAID levels
RAID, or redundant array of independent disks, is a method of grouping individual physical drives together to form one bigger drive called a RAID set. Because the server can simultaneously access more spindles (drives) for reads and writes, performance is improved when data is accessed from a “RAID-ed” drive.
The various ways in which data is grouped across drives is called the RAID level. Each RAID level is denoted by a number following the word RAID. The most common levels are RAID 0, RAID 1, RAID 5 and RAID 6. The RAID level is generally determined by the requirements of the applications running on the server. RAID 0 is the fastest, RAID 1 is the most reliable and RAID 5 is considered a good combination of both.
Combinations of RAID levels may be used together for additional data protection. For example, a user can create two RAID 0 sets, and then combine the RAID 0 sets into a RAID 1 set. This essentially provides the performance benefits of RAID 0 with the availability benefits of RAID 1.
It is important to note that not all storage array or RAID controller card vendors support all the various RAID levels. Before purchasing a storage system or RAID controller you should determine your organization’s specific needs relative to data protection and recovery, and application performance.
How RAID 1 compares to other RAID levels
RAID 1 requires a minimum of two physical drives, as data is written simultaneously to two places. The drives are essentially mirror images of each other, so if one drive fails, the other one can take over and provide access to the data that’s stored on that drive. Mirroring drives is good for very fast read operations, but it's slower when writing to the drives since the data needs to be written to two locations.
While RAID 1 drive mirroring provides comprehensive data protection that can be a lifesaver in the event of disaster that disables the primary drive, that level of data protection comes at a cost of requiring 100% drive capacity overhead. That means if you use drive mirroring RAID 1 on an array you have to double the amount of installed drive capacity—half to accommodate active storage and half for the mirrored copy. Because that amount of overhead capacity can be very expensive, very few enterprise-class storage systems use RAID 1, opting for more space-conserving RAID 5 or RAID 6 levels.
RAID 1 and RAID 10 are both mirroring technologies that use half of their available drives for data, but the main difference between them is the number of available drives. While RAID 1 uses two drives, RAID 10 involves at least four drives and combines the mirroring of RAID 1 with the striping of RAID 0.
RAID 1 vs. RAID 5 comes down to what's more important in terms of performance and cost. A RAID 1 drive pair can handle twice the amount of reads of a standalone disk drive. The main drawback of RAID 1 is that for the required disk space, the amount of capacity will double (e.g., to store 100 TB of data requires 200 TB of capacity). With RAID 5, if a user buys five disks, four-fifths of the capacity is usable, so much less additional capacity is needed. However, RAID 5 and RAID 6 cannot recover as quickly as a RAID 1 implementation. While RAID 1 may be able to offer a near instantaneous recovery time objective (RTO), RAID 5 and RAID 6 need time—considerable time for large capacities—to rebuild the data across their arrays.