One of the most expensive aspects of data protection is the cost of the storage that holds all of the copies of data. Backup storage capacity can be 10X or more the capacity of primary storage. Backup storage has a ferocious appetite eating IT budget and data center floor space.
While data efficiency features like compression and deduplication can lower backup costs and the consumption of data center floor space, it can’t hold back the data deluge. Many IT planners are looking to the cloud for rescue, but they may be disappointed, as legacy backup solutions don’t always fully exploit cloud capabilities.
Cloud Mirror or Archive?
Most enterprise backup solutions will use on-premises storage as the primary recovery source. It makes sense since most recoveries come from the most recent backup. All other backups typically serve a need to retain data. The problem is the typical enterprise backup application must have 100% of its data on-premises and can only make copies of that data to the cloud. In other words, they don’t have a groom function where they can move old backups to the cloud. These solutions use the cloud as a mirror of their on-premises storage.
A true cloud approach will allow archiving of older backups to cloud storage, keeping on-premises backup storage costs to a minimum. An on-premises copy only needs to contain the last 90 days or so of data, and can serve 99% of recoveries from that timeframe. A backup solution designed to archive to the cloud should be able to reduce on-premises storage by 8X or more.
Most legacy data protection solutions can leverage deduplication and compression to minimize the amount of data that gets transferred to the cloud. These solutions though, treat cloud storage as the end point. For the enterprise, this treatment is wasteful since most enterprises have multiple sites. If the cloud is just a final reciprocal target, no comparison is made with data from the other sites. Another level of deduplication, done in the cloud, comparing data between sites could drive down cloud storage consumption costs by another 3X or more. Since customers pay for cloud storage every month, the savings for the deeper level of deduplication has a compounding payoff.
Most legacy solutions also ignore the fact that most cloud providers have multiple tiers of storage. Amazon, for example (as we discuss in our article “Understanding The AWS Storage Portfolio”) has three tiers. The backup application should leverage all of these tiers but most legacy programs only use the middle tier, S3. Backup applications should use the EC2 Amazon storage tiers during a disaster recovery so the now cloud-hosted applications get high performing storage. Long-term storage of old backup data should use the Amazon Glacier tier to further drive costs down. Only using the middle tier means the customer is paying a premium for inactive data and not getting the performance they need during a recovery.
IT Planners need to make sure they peel back what a vendor truly delivers when they say, “cloud support.” IT planners should look for a solution that will limit on-premises storage costs AND optimize the use of storage capacity in the cloud. Leveraging both aspects of these efficiencies allows them to make cloud storage very price competitive.
To learn more about true cloud storage support, check out our on demand webinar, “The Best VMware Public Cloud Backup & DR Options.”