5 Years of the Enterprise Archive – massive scalability and cost savings
I’m Louis Imershein, Director of Product Management at Permabit Technology Corporation. When I joined Permabit over 5 years ago, we were a company focused on storage solutions for compliance and governance of information. Right around that time, we began to understand that the typical enterprise storage growth rates of 50-100% per year were not sustainable with current archive technology. Permabit embarked on a new course expanding beyond its roots in disk-based storage for compliance. We began to extend our existing technology to address a new type of storage we called the “enterprise archive”. Purpose built for the long-term retention of static information, an “enterprise archive” would address the problems associated with storage bloat in the coming years by being massively scalable, cost effective, 24×7 available and securely protected. Looking at 2007 growth rates, storage industry analyst George Crump observed that a 100 TB archive, could easily grow to over 3 PB by 2012 – a good indication we were on the right track! To manage such systems with then-current technology, and given ever increasing IT budget pressure, would be impossible. Something had to change!
From the very beginning it was obvious that supporting PBs of data would require a grid-based storage architecture. Storage grids deliver three essential aspects for an “enterprise archive”. First, a grid design (one consisting of multiple modular components) allows the system to scale to hundreds of PBs. Second, grids enable the system to scale both performance and capacity granularly while remaining online and available. Third, grid technologies eliminate the overhead associated with provisioning, and as a result eliminate 60% of the management tasks associated with storage administration.
In addition, Permabit understood that any solution designed to address storage bloat must leverage advanced deduplication techniques. We witnessed firsthand at our compliance customer sites, that islands of data were sometimes residing within entirely different parts of the infrastructure. A deduplication technology that only scaled to 16 TB volumes would be unable to save significant space from the vast pools of data spread throughout the enterprise.
We’ll talk about availability and secure protection in my next post…