Briefing Note: Introducing infinite io
There is little argument among IT professionals that at least 80% of data is never accessed again after creation. In fact, Storage Switzerland finds that three months after a one week creation-edit cycle as much as 95% of files may never be accessed again. However, there is a chance that 1% of that 95% will be accessed in the future. The problem is that there is no way of knowing what files will be in that 1%. As a result, a large chunk of the IT budget goes to storing, protecting and preserving all 95% of this data. Inactive data is the elephant in the data center; everyone knows it is there, but no one knows what to do about it. The purpose of infinite io’s network based storage controller is to get the elephant out of the data center.
A Whiteboard Full of Failures
The truth is that the disproportionate amount of inactive data has been a problem for years but the level of disproportion continues to grow, and there is no end in sight. It is being made worse thanks to machine-generated data and the very real potential to monetize old data. There have been plenty of attempts to address the problem of inactive data; hierarchal storage management (HSM), information lifecycle management (ILM) and most recently file virtualization.
By and large they have all failed. The first challenge that these solutions faced was that the capacity problem itself was not significant enough. In 1999, 1TB of storage seemed like a lot, now most laptops have a TB of storage. In 2015 inactive data stores are commonly in the dozens of TBs if not the hundreds. The second challenge these solutions faced was technological; the capabilities of the underlying hardware and software were not ready. The third challenge was the destination infrastructure; the secondary storage all had to be on-site. The organization saved little in terms of data center footprint and power costs.
Why Didn’t The Cloud Save Us?
Cloud storage is supposed to be the ideal archive, but it only addresses the destination infrastructure challenges, floor space and power. It also adds challenges. First, most cloud storage uses a different protocol than the data center’s NFS and CIFS. It also creates a latency challenge; this data still needs to be transferred efficiently to the cloud. Cloud Gateways helped by translating NFS and CIFS protocols to something that the cloud can support. They don’t solve the transfer speed issues however. They also don’t help identify the data that qualifies for being archived to the cloud.
Introducing infinite io’s Network-Based Storage Controller
The infinite io network-based storage controller is a layer–7 transparent proxy that sits in front of a data center’s network based storage. It uses deep packet inspection to manage and respond to file activity. The storage controller or the existing NAS resource store the active data. Data migration policies defined by IT identify inactive data, converts it into objects, compresses it, encrypts it and then migrates it a cloud-based object store.
Any of the migrated files appear as if they are still located on primary storage. If a user accesses a file, it is restored from the cloud back to primary storage until it prequalifies for one of the migration policies. This movement is all done without administrator involvement, allowing them to focus on more pressing matters.
Most migration tools tend to concentrate on file age. infinite io policies can be created to control migration on far more that just file activity. In addition to last access date, policies can be set based on any combination of file type, user ID, wildcards and other variables.
The return on investment potential of a solution like infinite io can go far beyond the reduction in storage purchases. Adopters of these solutions should see a reduction in data center floor space, power, and cooling cost as well as a dramatic improvement in the data protection process. Consider a backup job that now has 95% less data to protect. In the past, these compelling ROIs have been overshadowed by the cost and complexity of the solution. infinite io has developed a turnkey solution that should lower the cost and virtually eliminate the complexity.