As data volumes grow, there's a bigger problem meeting
recovery time objectives (RTO). We depend on our data, and we have set recovery times for our systems that are zero (no downtime) in some cases, though most companies set recovery times from eight to 24 hours. Whatever the case may be, we typically will not be able to meet our RTOs because there is simply so much data to be restored. This is especially true with file servers, and clients are regularly telling us that it takes a full weekend to do a full
backup of their file server using traditional backup methods, like tape. Now, if it takes a weekend to back up a file server, it will simply not be restored in 24 hours -- probably two to three times longer -- and the RTOs are not met. Traditional tape restorations are also hampered by media reliability issues, offsite storage costs and retrieval delays. There is also a lack of prioritization in determining what data to restore first. Many companies respond to this by investing more money in faster backup platforms (e.g., disk-to-disk
Remember that backups also tend to multiply the volume of data that you're trying to protect. Not only do you need several full backups over time, but you must also contend with numerous incremental backups. For example, if you make weekly full backups and keep each backup for a month, you'll have four full backups on hand at any given time, not counting incremental backups. As your corporate data grows out of control, that spiraling volume is simply multiplied in your backups. The time and tape storage costs can become significant. This is a case where data deduplication technology can play an important role in backup, though it will never replace sound retention policies and decision making.
Listen to the Unstructured data FAQ audiocast.
Go to the beginning of the Unstructured Data FAQ Guide.