Enterprise Storage is not just a Label

I have many anecdotes around the topic of Enterprise Storage, but the conversations in the past 2 weeks made it important for me to share this.

Enterprise Storage is …

Amusing, painful, angry

I get riled up whenever people do not want to be educated about Enterprise Storage. Here are a few that happened in the last 2 weeks.

[ Story #1 ]

A guy was building his own storage for cryptocurrency. He was informed by his supplier that the RAID card was enterprise, and he could get the best performance using “Enterprise” RAID-0.

• Well, “Enterprise” RAID-0 volume crashed, and he lost all data. Painfully, he said he lost a hefty sum financially

[ Story #2 ]

A media company complained about the reliability of previous storage vendor. The GM was shopping around and was told that there are “Enterprise” SATA drives and the reliability is as good, if not better than SAS drives.

• The company wanted a fully reliable Enterprise Storage system with 99.999% availability, and yet the SATA interface was not meant to build a more highly reliable enterprise storage. The GM insisted to use “Enterprise” SATA drives for his “enterprise” storage system instead of SAS.  

[ Story #3 ]

An IT admin of a manufacturing company claimed that they had an “Enterprise Storage” system for a few years, and could not figure out why his hard disk drives would die every 12-15 months.

• He figured out that the drives supplied by his vendor were consumer SATA drives, even though he was told it was an “Enterprise Storage” system when he bought the system.

Mindset and Attitude

The experiences from the 3 stories are the anecdotal evidences why “Enterprise Storage” should not be Label. Marketing is a wonderful thing but it has been known to shed truth in many different ways. Thus, we cannot just accept that it is “enterprise-capable” and “enterprise-ready” just because the storage vendors or the cloud service provider said so.

For example, an Enterprise SATA drive will read MTBF (mean time between failure) of 2.5 million hours and AFR (annualized failure rate) of 0.35% to 0.45%. Yet the reliability and uptime of the business should not hinge upon the storage vendor’s claim that this Enterprise SATA is “Enterprise”. The operations of a business rely on many different factors, including the storage networking, the compute systems, the applications and the entire ecosystem that serves the business.

Thus, when a cloud storage vendor claim that their object storage has 99.999999999% data durability does not translate to your business uptime of 11 x 9s availability.

We must assemble our thoughts and common sense that Enterprise Storage must be a mindset, not based on a marketing pitch. It must be an attitude that we constantly question to ensure that our business (depending on the cost of operational downtime) can withstand outages, and more importantly, resilient to operational failures.

A Guidebook to Enterprise Storage mindset

I have developed this enterprise mindset framework almost 22 years ago. I got the idea of creating the framework when I was reading Jon Toigo‘s book “The Holy Grail Data Storage Management“.  This hard cover book is a prized asset in my large collection of computer (yes, plenty of storage technology) books.

The Holy Grail of Data Storage Management

The framework is known to me as the acronym A.P.P.A.R.M.S.C.

• Availability
• Protection
• Performance
• Accessibility
• Recovery
• Management
• Security
• Compliance

Each data point in my Enterprise Storage framework is self-explanatory. But I have moulded a set of “questions” and “considerations” for each of the points, and assign its applicability to the technology and the solution pitched to me. Then I link and match these “Enterprise Storage” features to the demands and the requirements of the operations and the business that it serves.

Every organization is different, and along with this framework, I also seek to understand the workflow (data pipeline) and the data life cycle involved in the operations of the business. Furthermore, it is also important to understand the applications within that operational ecosystem, and gather the storage access behaviours of those applications and workloads (I call this A&W, not the root beer) along the data pipeline. I have written this in the past, and this diagram below (I have others for other industries) was what I pitched as training notes to some previous end users and resellers many years ago.

The data pipeline for storage management in Oil & Gas Exploration and Production

From an enterprise storage and even cloud storage, you can find the patterns of the A.P.P.A.R.M.S.C. that can be applied to build the Enterprise mindset. As this is developed and matured, it becomes an attitude.

Think Enterprise always

There is no fool proof forever-available or forever-performing Enterprise Storage or global cloud storage service. At some point, anything could break down. Just last week, Akamai (not storage related) took down a big part of the Internet. This incident, as with many others before, spells out our business dependency on a service that impacted data availability to serve our business. It was not directly linked to storage but the path to access the data was made unavailable for several hours.

While what I have written here may not be in vogue, and I have been called a storage dinosaur before, I believe there is an ounce of wisdom that transcend beyond a marketing label that says “Enterprise Storage”. This attitude has served me well. I hope this helps you too.