Recovery Management Solutions Arrive
Snapshot copy presents a consistent point-in-time view of changing data. There are many variations of snapshot copy. With snapshot copy, write operations are saved in a separate area of disk storage reserved for snapshot activity. Here the old value of the affected area or block can be saved in case the new block(s) is corrupted or to permit a fuzzy data image that can be used for a recovery. Snapshots provide data protection from intrusion and data corruption, but not from a device failure that contains the source copy of data.
CDP (Continuous Data Protection) is like a movie whereas snapshot copy is like a series of still images. CDP creates a journal copy that is a sequential time-stamped history of write events. Journals are typically kept for two to four days covering the period of maximum likelihood for a data recovery action. Journals are good for protecting from intrusion and data corruption, enabling restores to go back to a point in time before the corruption occurred. Sparse journaling is available to journal writes for only the more critical tasks.
Disk mirroring is implemented using either local or remote disk drives for all copies. Mirrors protect data against a disk hardware failure but if the data is corrupt on one drive, it will be corrupt on the other also. Mirrors are maintained from either synchronous or asynchronous write operations in two or more places. Mirroring eliminates the backup window, but doubles the amount of disk storage and the associated operating costs. Mirroring should always be accompanied by other data protection schemes, permitting a recovery from clean data that existed before the corruption occurred.
Storage administrators must choose from either asynchronous or synchronous mirroring. In synchronous mirroring, both the source and the target devices must acknowledge the write is completed before the next write can occur. This can degrade application performance, but keeps the mirrored elements synchronized.
For asynchronous mirroring, the source and target devices do not have to synchronize their writes and the writes occur independently. Asynchronous mirroring is often used to replicate data to distant locations.
A new technique called deduplication is quickly gaining interest. Deduplication changes the economics of storage and reduces recovery times for the disk-to-disk backup. Deduplication segments the incoming data stream usually into smaller elements than files or blocks in some cases, uniquely identifies the data segments, and then compares them to segments previously stored. If an incoming data segment has already been stored, a hash code derived pointer is created for it. If the segment is deemed to be unique, it is then further compressed with conventional algorithms for an average 2:1 size reduction, and stored to disk designated for deduplication storage. For the backup application, deduplication has shown significant reductions in backup time and the amount of disk storage for the backup application. It's likely that deduplication will extend to other applications that have repetitive data movement patterns in the future.
Replication provides an executable image of data to use if the primary copy is unavailable. Like a series of still images, replicated copies are complete data images that are taken at specified points in time. This represents the most complete method to protect data and should accompany any mirroring implementation. The number of copies maintained over time increases disk costs. Nonetheless replication can allow recovery operations to occur by going back to a point in time before the error occurred.
Source: Horison Information Strategies: Storage Navigator
© 2005 Horison
|
