In a previous blog post we made a virtual journey from a traditional focus on IT products towards an emerging Brave New World of IT services, describing in some detail the benefits this paradigm shift offers to consumers. Somewhere on this journey, we crossed an important bridge, where our attention shifted from “HOW does this product work?” to “WHAT does this service provide?” That may be a small step for our narrative, but it is a formidable obstacle for the legions of technical talent that build and ship discreet IT products.
In a subsequent installment, we made the distinction between “Service” and “Utility”, suggesting that it is possible to marry the benefits of consuming a service with the reality of building and shipping tangible products. Examples abound in the real world, and, ultimately, aren’t services – transportation, IT, or otherwise – built and delivered using tangible products?
Today, we’ll imagine that we are a traditional vendor of IT products, faced with the challenge to meet our customers’ desire to realize the benefits of “Service” consumption. Keen observers will smugly note that this is not completely unlike EMC’s predicament. Thus, let’s map out how we may bring some “Service” consumption benefits to the business of building and shipping the industry’s leading storage arrays.
“Service” consumption benefits include:
- A focus on WHAT the consumer gets, not HOW the service works,
- Reliable delivery (with effective monitoring) of agreed-upon QoS levels,
- Ultimate agility (ease, speed, disruption, cost, …) for any service changes,
- Pay-as-you-go licensing/pricing, with effective metering of real usage
- Fully transparent TCO with (ideally) 100% opex / 0% capex.
This list is a rich mix of demanding technology and go-to-market challenges. Each line item fully justifies a separate in-depth exploration, but – as a true technologist – I am tempted to somewhat gloss over (4) and (5) assuming that – modulo the “effective metering” capability in (4) – they can effectively be addressed under the “Flexible Consumption Model” umbrella discussed previously. Hence, going forward, we’ll focus this narrative on (1)-(3), plus the “effective metering” listed under (4).
Let us undertake a mental excursion in which a diner walks into a restaurant in search of dinner. The diner is greeted and seated, usually with a glass of ice-water, as the waiter delivers the restaurant’s menu, including (usually verbally recounted) daily specials. The ice-water is not germane to our discussion, but the menu is. The diner examines the menu, picks one of the available choices, and puts in her order. Dinner arrives a short while later, is consumed with gusto, and the diner exits the stage – hopefully fully satiated – after paying for her meal.
Let’s translate this allegory: A service consumer (diner) approaches a service provider (restaurant), uses a service access point (waiter) to examine a service catalog (menu) and select a desired service option (dinner). The service provider (restaurant) delivers the selected service option (dinner) via the service access point (waiter), and the happy service consumer (dinner) exits the stage after paying for the service.
Most often, the menu (i.e. Service Catalog) is the most integral part of the experience. While in some elite restaurants, a celebrity chef may be the star, most dining experiences don’t demand best of breed. The same is true of the service, as long as it is acceptable. The menu is the ultimate statement of WHAT the restaurant (i.e. Service Provider) offers, hence WHAT the diner (i.e. Service Consumer) eventually gets.
This is the central point: nowhere on the menu does it say what type of pots and pans are used in the kitchen; or how hot the gas stove and oven is running; or how many dishwashers the restaurant employs; or any specifics or details on HOW the restaurant is equipped, organized and managed (i.e. service is built). The menu (service catalog) and the waiter (the service access point) are the primary – if not only – interfaces between the consumer and the provider..
Back in our storage domain, the contemporary consumer of the services provided by a storage array is usually presented with a bewildering spectrum of technology details, e.g. RAID levels, consistency groups, cache partitioning, multi-pathing, a/sync replication, etc… All are very valuable architectural or product features, but not the things a metaphorical storage “diner” is truly interested in when looking for “dinner”.
Which really, REALLY begs the question: “What IS a storage consumer interested in”? Or, to put it another way – if not RAID levels and cache partitioning, what should a menu of storage services consist of?
Next Month: The anatomy of a (storage) Service Catalog.