Time To Combine Backup And Archives?

Archives are not backups. I, and many others, have said it a million times. We have preached the idea of creating an archive process that is separate from the backup process with potentially separate storage areas. With the ever-advancing capabilities of disk to optimize storage and tape to economize it, as well as software to better bridge the gap, it may be time to consider merging backup and archive into a single process.

The concept of having an archive process that is separate from the backup process used to make sense. An archive ideally should be for a single, final copy of data that you need to store for a long time and potentially for legal reasons maintain a chain of custody. When you need it back, the ability to find the data is probably more important than the ability to recover it quickly. Backup, on the other hand, should be for the most recent copies of data that you are going to need recovered relatively quickly. In most cases you are going to know exactly where the most recent backup is stored.

There are problems with archive and backup being separate processes. First, there is overhead in having to manage and maintain two separate processes, especially since you are probably using separate hardware and software to accomplish that task. Second, many organizations don't have sophisticated the archive needs motivated by regulatory demands that large enterprises do. Finally, many organizations simply don't have the resources to support a separate archive and the reality is that despite our complaining they count on backups to fulfill that role anyway.

What is needed then is a simpler archive process powered by technology that supports that simplicity. Simply using a standard disk backup device for storing long-term data may take that simplicity too far. As we discussed in our article "When Does Backup Archiving Make Sense" disk-based appliances optimized for the archive process, but without all the features and expense of an "enterprise archive," is what the IT manager should be looking for. These products can deliver the basics that most businesses need without the complexity that the more feature-rich enterprise products have. A key enabler of simpler archive hardware is going to be smarter software. There seems to be two approaches available to users today.

Two of the key ingredients for archiving is searching and location tracking. Traditionally, backup applications were not very good at this since they needed to keep backup metadata small to be able to maintain good backup performance. However, an increasing number of backup applications now are improving on their ability to maintain very large indexes while delivering very fast search results and not impacting the backup process. Some have even added the ability to do contextual searching.

If your backup application does not support large indexes, or if you need something even more powerful, there are third-party systems that can be added to the backup process to provide backup-software-independent contextual searching as we discussed in a recent article.

Another option is to use the capabilities set forth by the Active Archive Alliance and use software that makes tape and multiple tiers of disk act as one. Data can be moved to the archive as it ages or the archive can be the host for the data itself. In other words change your archive into a fileserver and then let the software manage where it resides.

The technology to make archiving and backup a single process while still maintaining the needed capabilities of each is now a reality and there are multiple software options to get there. Technology has certainly marched on since we pronounced that backup is not archive and the time may be right to consider merging the processes or adding intelligence to your backup process.

Follow Storage Switzerland on Twitter George Crump is lead analyst of Storage Switzerland, an IT analyst firm focused on the storage and virtualization segments. Storage Switzerland's disclosure statement.