Active Archive: Object Storage Matures into Hierarchical Storage Management
The key to introducing new technology is to make the technology easy to use, deploy and manage. As the new technology begins to take hold the users decide what features will come next, this is the maturity curve. Maturity development starts with performance, continuity and resiliency. These are the key factors in data management. As the technology continues to mature, admins demand better reporting and visibility. This is when the business leaders begin to take notice, all of a sudden there is a cost of doing business that needs to be reviewed for savings.
Object storage was developed with this exact maturity map, and today object storage is the fundamental basis for the largest data management infrastructures in the world, cloud storage like AWS™ and Azure™. Object storage has also reached the point in maturity, where business leaders are looking to not only increase revenue, but also to reduce expense associated with long-term storage of digital data. This means object storage either has to embrace Hierarchical Storage Management (HSM) or expect that new technologies will supplant it in the market as the active archive of choice. The great part of this decision is HSM is mature enough to be integrated through orchestration and developed into the fundamental specifications of object storage. The question is where to store data that is rarely or never touched as it resides in the active archive? Tape, of course!
That is right, tape is leading the industry with solutions that store object data on the lowest cost, most bit-reliable storage medium. This is a story as old as digital time, all data that needs to be retained for any significant amount of time, glacial data, ends up on tape (#alldataendsupontape). The flexibility of tape is that the technology is a fundamental storage medium controlled by Software Defined Storage (SDS) and deployed as SDS long before SDS was popular. As long as data is in bit structure, tape can retain it.
There are plenty of solutions in the market from the top tape companies that support storage of object data on tape, the solutions are not the critical part of the discussion, the economics of this transition is the big controversy. Object storage purists will argue that object storage on disk is just as inexpensive, and easier to manage than deploying tape in the active archive. Rather than argue the point, let’s make one consideration, if the purist statement was true, why in the world would the top cloud storage providers deploy tape in their active archives? Because even at the economic scales of Amazon AWS or Microsoft Azure, tape is 4 to 5 times less expensive to deploy and manage than HDD infrastructure. This calculation is without taking into account the ability to meet or exceed Power Usage Effectiveness (PUE) metrics.
Since this is an economics of scale decision, what is the real compare of running a true active archive, based object storage infrastructure that includes tape? According to conservative estimates tape reduces the overall cost of glacial storage by an average of 56%, when compared to HDD object storage alone. This is based on five years of retention with 5 Petabytes of data, 50% of which is glacial data.
Active archives utilizing object storage and tape are the future of data retention across both enterprises and cloud providers. As the two markets merge in deployments, there will not be a question as to “how and why do we want to HSM data in our active archives”, it will simply be “how much can we save.” Tape is ready for accepting this and all data. It is estimated that tape is currently operating in over 150 countries around the world.
There is no reason to delay setting up the infrastructure to optimize the object storage infrastructure with tape. The sooner your organization starts the journey, the more likely you will be to recognize success in business line outcomes.