Primary storage has two key functions, to deliver data as fast as possible to the applications and users requesting it, and to maintain data accessibility in the event of a hardware failure. Primary storage vendors have attempted to expand their role beyond those two key functions, leading some to claim that they are self-protecting primary storage and can displace the backup process.
There is a standard that all data protection processes must meet, the 3-2-1 rule. It states that a solid data protection strategy must have at least three copies of data, on two different pieces of media, with one copy off-site. Primary storage vendors will claim that a combination of snapshots and replication enables them to meet these challenges.
Dissecting the 3-2-1 rule
The first part of the 3-2-1 rule is three copies of data. Note, this is not three versions of data, which is essentially what a snapshot provides. Three copies means, three independent and complete copies of the same data. Snapshots are totally dependent on primary storage. If it fails, so do all the snapshots. In theory part of the “three” could be met by storing two complete copies on the same storage system but that is not best practice. The spirit of the three is three independent copies on three separate storage systems.
For a primary storage system to meet the “Rule of Three” requirement it needs to take a snapshot and replicate that snapshot to two separate storage systems, which also enables it to meet the “Rule of Two.” Most primary storage systems have a snapshot and replication feature, and they can even have different snapshot schedules on each system, giving them multi-versioning. If the storage system then replicates data off-site, they then meet the “Rule of Three” and the “Rule of 1”, which is one copy off-site.
While many primary storage systems can meet the basics of the 3-2-1 rule, they do so with a lot of problems. The first is the expense. Most primary storage systems can only replicate to another storage system identical to itself. In the case of an all-flash array, replicating to an identical system means the organization has to buy three flash arrays.
The second problem is capacity. Most organizations leverage the backup process to restore previous versions of a file and to retain certain versions for a period of time. While most primary storage systems claim “unlimited” snapshots, in most cases maintaining a series of snapshots for years will eventually impact performance. It also consumes disk space since those changes need to be tracked until the snapshots eventually expire.
What IT Needs
If IT is going to invest in self-protecting primary storage it needs to make sure that the solution does not require the second copy of data to be stored on-premises. It also needs to avoid storing snapshot versions on primary storage.
In our on demand webinar, “How to Design Self-Protecting Storage and Gain Backup Independence”, Storage Switzerland and ClearSky Data describe an architecture that uses the public cloud in combination with a regional cloud to provide a cost effective and more simplistic approach to having primary storage meet the 3-2-1 rule. Three copies are kept, one on-premises, one in the regional cloud and one in the public cloud. Snapshots are maintained in the regional cloud and public cloud. The regional cloud and public cloud also provide an automatic off-site function. The result is not only better data protection but also a more affordable solution that reduces on-premises investment and footprint, while leveraging the regional and public cloud for long term data retention.
Interested in learning more? Access our on demand webinar with ClearSky Data, “How to Design Self-Protecting Storage and Gain Backup Independence”.