How to Justify Spending $500,000 a Month on Archival Storage
The lead article in the March 2009 issue of Storage magazine, "Quality Awards for Midrange Arrays", finds that a remarkable 82% of respondents in the survey would make the same storage purchasing decision that they made in the past. But are these impressive numbers a sign, as the author of this article suggests, that users are truly satisfied with their current storage system? Or is it a sign of a deeper, more systemic problem in storage where users resist changing their storage system even when they are presented with a solution that is obviously better and/or more economical than the one they currently use?
This point was driven home to me in a recent conversation that I had with a systems engineer who was discussing with one organization how it was currently storing its archival data. It turns out this company was keeping all of its archival data on a primary storage system and paying its current storage system vendor $500,000 a month for the privilege of doing so.
Eventually the course of the conversation turned and the organization asked the engineer, "How long would it take him to show an ROI if it replaced its current storage system with a Permabit Enterprise Archive archival storage system?" On hearing the question, he paused, thought about it and then replied (partly in jest), "About 8 hours." It's when I hear anecdotal stories like this that it begs the question, "Why are organizations so comfortable - and how can they justify - spending millions of dollars a year on storage solutions that are obviously overkill for their application data requirements?"
The rationale behind making such buying decisions may seem illogical in the light of day but it goes back to what motivated organizations to buy these primary storage systems in the first place. Years ago storage administrators worked their tails off to convince management that spending extra money on primary storage systems was a necessity for their production application data. Once everyone saw the benefits of primary storage (high availability, reliability and performance), they wanted to store as much data as they could possibly justify on these systems because they believed other storage systems didn't meet their requirements.
But in the last 5 - 10 years, everything changed. Not only did the amount of data that enterprise organizations need to keep on disk increase, but the nature of the data itself changed. While obviously some production data still requires primary storage systems with their high end features, a greater percentage today does not. For example, up to 60-80% of data on primary storage has been identified as not needed in the transactional performance storage tier. But where organizations can get trapped is that the same mindset that prevented them from adopting primary storage in the first place is now keeping them from replacing it with a more suitable storage solution.
In one respect, one can understand why this organization kept all of its archival data on primary storage. After all, until the current economic downturn hit, what was the reward for these individuals to move this data to a more suitable, cost-effective storage system? In truth, there was not much of an incentive. If they did purchase a lower cost solution, they might save some money but with their company awash in cash, who would notice or care? And then if, God forbid, something did go wrong, all fingers are pointing at them because they are the one who recommended the alternative solutions. So $500,000 a month to avoid being blamed for data loss becomes more logical in a certain twisted sort of way. I guess?
There is nothing wrong with being loyal to your storage vendor who has taken care of you over the years, but there comes a point where it reaches a point of insanity. I believe that some organizations have now reached that point. When they can purchase archival storage solutions such as the Permabit Enterprise Archive for a fraction of the $500,000 price and still get all of the features that they need to meet their storage requirements, it's time to re-evaluate what storage you are buying and who you are buying it from. And with this economic climate, that may be just thing the thing that saves your job!
This point was driven home to me in a recent conversation that I had with a systems engineer who was discussing with one organization how it was currently storing its archival data. It turns out this company was keeping all of its archival data on a primary storage system and paying its current storage system vendor $500,000 a month for the privilege of doing so.
Eventually the course of the conversation turned and the organization asked the engineer, "How long would it take him to show an ROI if it replaced its current storage system with a Permabit Enterprise Archive archival storage system?" On hearing the question, he paused, thought about it and then replied (partly in jest), "About 8 hours." It's when I hear anecdotal stories like this that it begs the question, "Why are organizations so comfortable - and how can they justify - spending millions of dollars a year on storage solutions that are obviously overkill for their application data requirements?"
The rationale behind making such buying decisions may seem illogical in the light of day but it goes back to what motivated organizations to buy these primary storage systems in the first place. Years ago storage administrators worked their tails off to convince management that spending extra money on primary storage systems was a necessity for their production application data. Once everyone saw the benefits of primary storage (high availability, reliability and performance), they wanted to store as much data as they could possibly justify on these systems because they believed other storage systems didn't meet their requirements.
But in the last 5 - 10 years, everything changed. Not only did the amount of data that enterprise organizations need to keep on disk increase, but the nature of the data itself changed. While obviously some production data still requires primary storage systems with their high end features, a greater percentage today does not. For example, up to 60-80% of data on primary storage has been identified as not needed in the transactional performance storage tier. But where organizations can get trapped is that the same mindset that prevented them from adopting primary storage in the first place is now keeping them from replacing it with a more suitable storage solution.
In one respect, one can understand why this organization kept all of its archival data on primary storage. After all, until the current economic downturn hit, what was the reward for these individuals to move this data to a more suitable, cost-effective storage system? In truth, there was not much of an incentive. If they did purchase a lower cost solution, they might save some money but with their company awash in cash, who would notice or care? And then if, God forbid, something did go wrong, all fingers are pointing at them because they are the one who recommended the alternative solutions. So $500,000 a month to avoid being blamed for data loss becomes more logical in a certain twisted sort of way. I guess?
There is nothing wrong with being loyal to your storage vendor who has taken care of you over the years, but there comes a point where it reaches a point of insanity. I believe that some organizations have now reached that point. When they can purchase archival storage solutions such as the Permabit Enterprise Archive for a fraction of the $500,000 price and still get all of the features that they need to meet their storage requirements, it's time to re-evaluate what storage you are buying and who you are buying it from. And with this economic climate, that may be just thing the thing that saves your job!
Leave a comment