News Stay informed about the latest enterprise technology news and product updates.

Backups and archives require separate discussions

I’ve spent a lot of time talking to IT clients and vendors about archiving and backup, and these discussions made me realize there is a real problem with the terminology used today.

For IT clients, the terms have different meanings depending on the responsibilities of the person talking. Preconceptions (or misconceptions) give color to what motivates customers in managing information. The application owner or business owner sees backup as something for which IT is responsible for. At the same time, IT sees archiving as a possible impediment to success on because it could make it more difficult to access needed information.

Vendors approaches to backup and archiving are driven by their products. For most vendors, backup and archiving are usually combined and their messaging may cover both at the same time. This may not serve the vendor well because of the different customer perceptions and people served.

There are a few basics about the terminology that need to be understood along with some recommendations:

Backup is really about data protection. Data protection should be the top level message and is a continuum that includes replication and point-in-time copy (snapshot). Today, backup is an IT function where the backup group in IT serves the overall business – both applications and systems.

Archiving is really about information management. For the IT backup guys, it is just another form of backup and usually is thought of as backups that are being kept (retained backups) rather than part of a rotation. For the application owner, an archive is about moving some data and making it difficult (or delayed) to access. A storage administrator sees archiving as a migration between tiers and a way to reduce the primary capacity demand as part of capacity management.

The archiving discussion must be separate from data protection, although there is a data protection component in archiving. IT rarely takes initiatives to implement archiving practices (other than retained backups) for several reasons:

• Usually IT is not empowered to make decisions about application data from business owners. The idea that data can be made less accessible or deleted is not something IT people believe they have the authority to do.

• IT does not want to be wrong and cause an impact when it comes to making a decision about the data. The negatives outweigh the improvements that may be made by implementing an archiving strategy.

• The assignment for archiving in IT usually lands in the purview of the backup manager/administrator. Managing the backups is challenging and archiving is seen as moving individual elements such as files that are too fine-grained for the backup process.

The archiving practice needs to focus on the application and business owners who ultimately are responsible – both for the application use and the economic costs. The approach should be about moving data to a content repository that is appropriate for the diminished probability of access. The content repository is less costly, but must still be directly accessible by the application and the information (typically files) visible to the application owner. The content repository is not about files stored inside a backup format but as individual elements (files) that are, in the application owner’s terminology, online.

From the application owner’s perspective, IT is not involved in the access. There should still be a discussion about “deep archive” repository for data that is not expected to be needed again but cannot be deleted. Again, this is an application owner decision but the mechanics are implemented by IT.

When it comes to backup and archiving, terminology matters. There is context for usage and dependencies on who is involved in the discussion. Archiving must be considered in the context of the application. To counter the preconceptions, the discussion should be about application content repositories rather than an archive. The concept of a deep archive is still highly valuable. The archive discussion needs to be with the application owner. Backup needs to be put in the broader context of data protection and separated from archiving. This makes the discussion more relevant to those involved. It also makes it an easier discussion to have.

(Randy Kerns is Senior Strategist at Evaluator Group, an IT analyst firm).

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.