Since individual technologies seem to change at a rapid-fire pace, it’s shocking how long magnetic tape media has survived (first used in 1951 to record computer data, it outdates hard drives and is now approaching 61 years of use!). Although innovative new mediums (e.g., hard drives and solid-state storage) have exponentially increased speed and reduced the physical size of storage over the decades, cost and/or scaling issues left magnetic tapes as the logical disaster recovery choice for most organizations. Recent innovations in both storage mediums and performance enablers, however, may be the catalysts necessary to finally move beyond tape. This article will explore one alternative, electronic data vaulting and the use of virtual tape libraries, and compare its benefits and shortcomings to magnetic tape.
What Is Electronic Data Vaulting and How Does It Work?
Electronic data vaulting works by electronically transferring backup data from a primary site to a remote storage device using wide area network bandwidth. While this sounds similar to synchronous/asynchronous replication, it differs in several key ways:
- Typically, backup data for electronic vaulting is kept on high density, low cost hard drives, rather than expensive matching SAN arrays
- Rather than transferring every change to a remote site, de-duplication software is used to consolidate a batch of changes and reduce bandwidth requirements
- Typically, transfers to the remote site happen anywhere from hourly to daily, requiring far less bandwidth than replication of the same data.
Organizations also typically utilize a type of emulation software that models data storage and retrieval using “virtual tapes” within a virtual tape library. This structure enables organizations to easily integrate the solution into already existing backup strategies – in the eyes of the end servers, there is no change in how data is backed up, just in the media on which the data is stored.
Benefits of Electronic Data Vaulting
As data is written to electronic media instead of magnetic tape, the process works more quickly. In addition, electronic data vaulting can enable quicker backup and recovery by replicating only either byte or block level changes, which reduces the required backup window and enables more data to be backed up within the same timeframe. As electronic vaulting does not require real-time replication or replication during peak timeframes, the data pipe necessary to transfer data can be smaller, making it cheaper. Since tapes no longer have to be physically moved offsite, this removes the need for environmental controls and/or risk of environmental damage to the tapes during transit, one of several reasons tapes fail.
Also, when restoration is necessary, data can be immediately accessed via either the local or remote copy, reducing the time necessary to recall tapes from offsite storage. Individual files can be easily located via an electronic inventory, rather than loading and unloading tapes to restore the necessary data, which can save hours of technician time. Expanding to store additional data is typically as easy as adding additional disk drives, while tape can often require the purchase of additional tape drives, in addition to extra tapes.
Downsides to Electronic Data Vaulting
Though electronic vaulting may result in cost savings for some companies, it is not true for all organizations. The larger the amount of data, the less cost-effective electronic vaulting can be versus tape, since disk is a more expensive medium than tape, and the more data necessary for backup, the more tape infrastructure costs (the expensive part) can be spread across the cost per GB. In addition, depending on how much data must be sent within available backup windows and how far apart the primary data center and disk backup are, bandwidth can get extremely expensive. Each organization must assess their requirements and risk tolerance to determine whether the acceptable distance and recovery points make disk backup financially feasible.
In addition, a major potential downside of electronic vaulting is archiving. For organizations required to maintain electronic copies of records or files, solutions that apply delta changes to files essentially “erase” the previous instance of the file, as it can no longer easily be recreated to the point in time necessary to meet regulatory requirements. While delta logs might make it possible to “roll back” a file’s delta changes for a period of time, it’s not an easy process, and most delta file changes get “further consolidated” into specific periods (e.g., dailies into a weekly file, weeklies into a monthly file, monthlies into a quarterly file), which makes rollback beyond that point impossible. When you compare this to tape, where you can just ship specific tapes offsite for “permanent storage”, the cost/difficulties could outweigh the benefits for organizations where data archival is common practice.
It’s also important to note that not all backup software or storage arrays support the “de-duplication” process. Vendors have been able to design most new software/hardware to handle the application of only delta changes to files, but older software cannot always enable this. A lesser barrier but still important issue occurs with virtualization – though almost all virtualization software supports recovery from virtual disk, the process is not as automated as replication of the servers would be, which complicates the recovery process.
Factors to Consider When Evaluating Electronic Data Vaulting
Some organizations choose to outsource electronic vaulting to a third party provider rather than managing the data and secondary environment themselves. This solution prevents the need for alternate location environment and server maintenance/refresh, which can reduce internal complexities and overhead, though your organization will need to ensure it is comfortable with in-place security measures.
Whether your organization opts to internalize or outsource electronic vaulting, there are a number of factors that can reduce or increase costs. First, the size and frequency of data backup affects performance and required bandwidth, as data that must be transmitted regularly (i.e. during business hours) would require a smaller but dedicated pipe, while backing up data nightly must have a network pipe large enough to back up all data during non-peak hours. Setting up several data tiers can further reduce costs, as for example, critical data could be backed up regularly (e.g., 30 minute intervals), semi-critical data could be backed up every four hours, and non-critical data could be backed up nightly.
“The year ahead will see a major increase in cloud based recovery services (the front end of which begins with electronic data vaulting – or “Cloud Backup”) since they relieve customers of the need to manage backup tapes at their production sites and help them eliminate tape-based restores as their primary method of data recovery – all making the whole backup and recovery process more efficient and effective.”
– Dick Fordham, Director Marketing and Strategy, Recovery Point
While electronic vaulting may enable your data to be available offsite, you must keep in mind that it’s not like a hot site – in addition to the storage to which data is replicated, you must either have alternate equipment on hand or set up a strategy to acquire necessary equipment as quickly as would be necessary to support recovery objectives. While electronic data vaulting removes the “tape restore” time, you must still account for the time necessary to restore servers and point them to the appropriate data.
Overall, electronic data vaulting may not be the right fit for every organization. However, it is definitely worth the effort to assess pros and cons and determine which backup and recovery solution makes the most sense for your organization, rather than just going with the “tried and true” (especially considering the relatively high failure rate we observe with tape backups). When determining if this newer technology meets your organization’s needs, consider the following:
- All costs (direct and indirect) associated with your current solution
- Current and anticipated future data storage/backup needs
- Retention requirements
- Recovery requirements (how quickly data must be restored/useable)
- Risk tolerance for both distance between the primary and backup location and tolerance for involving external 3rd parties with maintaining data
- Bandwidth requirements
As with any technology, there’s no one-size fits all solution, but this one’s worth a second look, particularly if you’re able to put in the necessary analysis into your current costs, future needs, and risk adversity.
Avalution Consulting: Business Continuity Consulting