Permabits and Petabytes

June 25, 2008

Enterprise Archive Defined

Filed under: Jered Floyd — jeredfloyd @ 2:48 am

The principal storage market that Permabit addresses, with the aptly named Permabit Enterprise Archive appliance, is the area of enterprise archive. Analyst George Crump does a great job explaining this market and its requirements in his white paper The Enterprise Archive Defined, but I’ll take a crack at a brief summary here.

Enterprises today are faced with tremendously growing storage requirements. Any market analyst (or, better yet, scientifically conducted survey) will tell you that storage requirements in enterprise are growing between 40% and 60% yearly. This is driven by many factors (regulatory requirements, data mining, modernization of business, higher resolution media, etc.), but it all comes down to wanting and needing to store more stuff.

Most of this stuff goes right onto primary storage, at a cost of $30 to $50 a gigabyte. It also gets put into the backup stream, at an overall cost (considering how many copies are eventually made) of another few dollars per GB. At current growth rates, the average mid-size enterprise needs millions of dollars more primary storage per quarter, but IT budgets are certainly not keeping up.

Some businesses try shuffling this directly off to tape or optical as an off-line (or near-line) archive, at perhaps $3 to $5/GB, but such solutions just don’t meet the requirements. As George talks about in his paper, tape and optical libraries don’t offer the reliability and availability necessary for enterprise archive data — after all, this was data that was previously being kept on Tier One storage!

At Permabit, we think there’s a better solution. Our purpose-built grid architecture allows us to use of-the-shelf hardware components to create a super high-reliability disk archive, but at a much lower cost than other disk-based solutions. Using standard components also allows us to seamlessly take advantage of the ceaseless march of Moore’s law, providing higher performance and density without having to redesign our architecture. With technologies like RAIN-EC, which protects data across multiple nodes in the grid, we can deliver reliability that exceeds RAID 6. These allow us to reach storage costs of $5/GB or lower, and that’s just the start.

Permabit Enterprise Archive also integrates Scalable Data Reduction (SDR), which is our name for a combination of traditional compression methods and sub-file deduplication (which is rather like compression across files, if you think about it). By identifying common data across the entire system of perhaps hundreds of terabytes of data, we can deliver effective storage costs well below that of any other solution. Permabit’s SDR is unique in that it is optimized for archive data, where we have to work harder to find commonalities in data than in the more common backup/VTL case, where the basic system design provides certainty of around 25 copies of each piece of information. SDR also scales to petabytes of real physical disk, whereas other dedupe technologies top out at 30 or 40 TB (but claim that as petabytes…)

So, the short version of enterprise archive data is information that’s not being used in hundreds of transactions per second, requiring top-tier primary storage performance, but needs to be kept around reliability, with high availability, and inexpensively. We’ve developed Permabit Enterprise Archive to be the optimal place to put this data, whether it’s written there initially, or moved to us by a separate archiving application.

Note that “deduplication” is not a requirement for enterprise archive, but it’s part of our solution that allows us to deliver incredible cost savings. That’s because deduplication is a great technology, but it’s not a user feature… more on that next time.


Leave a Comment »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

%d bloggers like this: