Ever buy an inexpensive set of glassware only to have it frost up after a couple of cycles through the dishwasher? How about buying the lower priced kitchen appliance that appears to look like all of the rest, only to have the internal motor die shortly after the limited warranty expires? Unfortunately, no matter the analogy, we’ve all been there in one form or fashion. The good news is that most of these errant purchases typically only result in a little bit of frustration followed by a re-purchase of a higher quality alternative.
We all know that the axiom of “you get what you pay for” is certainly true in the tech world. Like the world of radial tires, there are low cost, mid-range, and high cost/high quality products in every nook and cranny in the tech space. As with every evolving technology, this axiom is proving to be true in the tech world, but with a slight twist.
Not too long ago, large tech companies like Microsoft, AWS, and Google, in an effort to rapidly scale out their public cloud solutions, began making the design and purchasing of cloud environments from their websites relatively easy. This meant the consumer, yes the consumers, could now design their own cloud infrastructures without any oversight…or expertise. Imagine building an infrastructure for a business critical application without any redundancy, backups, or DR? Just as problematic, we often see companies that have designed these types of environments without SLA’s or Support options.
As a Managed Service Provider (MSP), we have seen more of our share or rescue engagements from companies that thought they could just “spin up a server” and off to the races they’d go. Imagine your companies’ website going down because you had not designed it with redundancy and worse, did not think to add support resulting no support escalation path? What is missing with these tools within Microsoft Azure, AWS, and Google are logic rules that prompt the users to consider these options…and the consequences of not doing so. An unfortunate bi-product of enabling the consumer to design and purchase flimsy and inexpensive infrastructure environments is when there are issues, and there inevitably are, the big public cloud companies get the negative publicity. In reality, these environments when designed correctly carry the highest levels of uptime and are coupled with the excellent support paths.
So what happens if this scenario occurs within your organization? Scour the Internet looking for remedies to slow or underperforming systems. This can be dizzying and you will likely find little relief. Support services for the big public cloud companies do not include design and implementation consultation services, so that is really not an option. Fortunately, there are dozens of qualified Managed Service Providers s that can provide comprehensive approaches to designing a solution that meets your companies’ goals. Unfortunately, there are hundreds of “unqualified” firms that will gladly tout their infrastructure expertise as part of a larger holistic suite of offerings…think Platte River Associates from the Hillary Clinton email saga.
First and foremost, check the credentials of the firm. Are they SOC 2, Type 2 certified? Not only the data centers, but the company. This certification ensures that industry recognized Best Practices are being followed for design, security, support, etc. Next check the resumes of their engineering team. If they don’t carry the appropriate credentials, then it’s likely an expertise gap exists. Lastly, check references and inquire about the type and size of clients they have within their stable. Many firms will overstate their expertise or client “sweet spot” when chasing new opportunities.
Which brings us back to getting what you pay for. If you were going to buy a new car, and the dealer had a kiosk that enabled you to design your own engine for a discount, would you? Of course not…you would go with the car designed by their automotive experts. The same thought process should be applied when determining how to design an infrastructure for mission critical applications…eave it to the experts. A poorly designed system, although inexpensive from a monthly acquisition fee, usually will cost many times more in the long run when calculating slow performance, system downtime, loss of data, and employee frustration. Trust me, when the users are on the warpath, the inexpensive monthly cost of the infrastructure will be the last excuse they will want to hear.