| Backup may have been around forever, but it sure looks different these days.
Data backup is as old as computing itself. Ever since the first paper tapes and punch-card decks were duplicated, IT admins have recognized the need for multiple copies to guard against the unexpected. But even in the days of batch computing, accomplishing this task presented the ever-familiar challenges of resource constraints, limited time to accomplish the backup and managing multiple copies.
Until recently, the basic nature of backup has remained largely a batch-oriented tape management process. Magnetic tape, evolving for decades in terms of speed, capacity and reliability, was the core technology that enabled offsite and long-term storage of backed up data at an affordable price point.
The transition to LAN-based backup in the late '80s and early '90s brought economies of scale and improved the manageability of the backup process. No longer was it necessary to arm each server with a dedicated tape drive or enlist armies of tape jockeys to load/retrieve tapes.
With network-based backup came the rise of robotic tape libraries and commercial client/ server backup apps. Before then, backup was most commonly performed by system utilities that were often provided with the OS. Commercial apps like Legato (now EMC) NetWorker, OpenVision (now Symantec's Veritas) NetBackup and IBM ADSM (now IBM Tivoli Storage Manager) led the way and captured the lion's share--roughly 70%, according to research firm Gartner Inc.--of the enterprise backup market.
Readers of this magazine are well aware of the current technology revolution underway in the field of data protection and recovery. As disk (in its many guises) plays an increasingly prominent role in the process, the backup app is changing as never before, and it's important to consider its impact on recovery management.
The forces driving the underlying need for change in the backup realm include:
The end of backup software?
At the same time, the range of available options for data recovery expanded well beyond backup to the point that, for many organizations, restoring from tape has become their option of last resort.
This world of recovery possibilities, enabled by disk technology, is pushing backup apps into a new role. Backup apps are evolving as the central entity for managing the multiple means by which disk is used for recovery. This includes:
Snapshot management: Snapshots and split mirrors have long been used to facilitate the backup of key apps. Beyond improving backup, many environments snapshot throughout the day and retain snapshots. In most cases, they're managed independent of backup.
Replication: Like snapshots, replication has become an important recovery mechanism, particularly for disaster recovery, and needs to be included in any comprehensive recovery management capability.
Virtual tape libraries (VTLs) as intelligent disk: For large tape environments, tape emulation remains the simplest and least disruptive means of leveraging disk, and VTL vendors are working to enrich their feature sets beyond just emulation. However, functions performed outside of the awareness of the backup app (e.g., replication) can often create confusion when it comes to recovery and overall data recovery management.
Deduplication: Deduplication has created opportunity and confusion. It's a technology that can be performed at the client, server, within a NAS appliance or in a VTL. Backup applications must be able to accommodate any of those approaches and often more than one at a time.
Continuous data protection (CDP): Like deduplication, CDP is a technology that takes a number of forms. In some cases, it exists as a standalone technology, but its real value is as an integrated element of a tiered recovery strategy, and it should be managed as such.
Enhanced application recovery: In some cases, disk can be used to simplify and improve recovery capabilities for environments like VMware and apps such as Exchange and SharePoint with both volume-level and granular object recovery available from a single backup.
Products evolve with service features
Veritas NetBackup (with Version 6.5 and beyond) is becoming well-positioned as a multifaceted data recovery manager. Support for a wide range of snapshot technologies, and its OpenStorage API for integration of intelligent disk functionality, bode well for the future. Also adding to its versatility is software-based deduplication through the PureDisk option and support for the recently announced CDP-based RealTime Protection.
A key enhancement on the management side is the notion of Storage Lifecycle Policies that enable automated business-oriented management of the multiple copies of data that might exist in various forms. For example, it's possible to establish a policy that snapshots a file and then copies it to a deduplicated storage device and an offsite tape, each with a different retention time. The data movement and subsequent expiration is fully automated; in its latest incarnation, Storage Lifecycle Policies support a variety of snapshot technologies as well as CDP.
IBM Tivoli Storage Manager (TSM) can be considered the "granddaddy" of commercial disk-based backup products, as it was originally designed to support the hierarchical movement of storage across various media. TSM supports a wide range of disk functionality like snapshots within the IBM portfolio, as well as selected third-party support. With regard to management functionality, as a member of the Tivoli management portfolio, it offers a number of enhancements through related products (like Tivoli Storage Process Manager) to provide ITIL-based operational processes.
EMC has been extremely active in the data recovery realm with the acquisition of a range of companies dealing with backup, CDP and data deduplication. NetWorker has become an integration focal point for these and other technologies. The result is a rich set of support for snapshot technologies through a variety of PowerSnap modules, as well as support for EMC's Avamar deduplication backup product and RecoverPoint CDP offering. A dashboard management console lets users view and drill down through many of these components.
Looking beyond the three giants, one relative newcomer, CommVault, has managed to make inroads in enterprise data recovery. Taking advantage of not having to support the historic product legacy of its competition, its design focus has been on multifaceted data management and recovery; its Simpana suite integrates backup, archiving and snapshot, as well as keyword search and retrieval within a common management framework.
As more disk-based recovery alternatives are adopted, the coordinated management of multisource recovery will continue to increase in importance. A logical place to look for this to coalesce is within the backup application. It's important for organizations to understand their vendor's strength and direction in this area. The time has come to shed the traditional, compartmentalized view of backup and embrace a recovery-centric data management focus. To do otherwise would mean standing still.