Organisations buy storage infrastructure for one reason: meeting application service level objectives (SLOs). Applications look to storage for availability/accessibility, performance, and protection. While these functions may seem simple, a look at all of the different storage system and software offerings in the marketplace shows that it’s one of the most complex challenges for any data center.
Most storage service level discussions begin with availability and performance. To meet those SLOs, teams deploy multiple storage personalities and configurations – high-performance block storage or scale-out object storage or raw, low-cost IOPs storage, etc. Then they consider protection.
Protection is becoming exponentially more challenging to select and provision. Protection SLOs include a recovery point objective, a recovery time objective, version retention, and geographical redundancy. To try to meet the SLO, each storage array, hypervisor, and application offers multiple protection technologies (e.g., archival, backup, replication, clones, and snapshots). The result is a sprawling set of infrastructure configurations, which can be difficult and costly to manage, maintain, and adapt to the application environment.
That’s why, in 2014, the storage market will begin the quest for SLO-Defined Storage with a real dragon to be slain around data protection. The answer won’t be a one-size-fits-all product, but a solution that configures the appropriate data protection mechanisms when setting up the primary storage.
In 2014, customers will evaluate storage on how cost effectively their architectures deliver to their application SLOs across availability, performance and protection.
The battle cry will be: “I want to provision protected storage.”
About the author
Stephen Manley is Chief Technology Officer, Data Protection & Availability Division at EMC
This article first appeared here.