The adoption of cloud based and hybrid environments have data centers changing their strategy. Recently both Verizon and CenturyLink sold off portions of their data center business. With the rise of companies like Amazon Web Services, your data center will have to provide you with added value.
Most data centers are now selling a cloud environment they built to profit from their additional capacity. Whether traditional a traditional power, ping, and cage or a cloud environment, the right connectivity is crucial.
Some data centers were able to charge a premium for their ability to reduce latency through geographic proximity for certain industries. Today’s customers want to know how much per meg for bandwidth, where’s the nearest IP router, and how many diverse paths do you have? All of these are essential but with all of the carriers available at the data centers, how do you know you’re getting the best price, on the most diverse route, from the provider who actually owns the right to service the wire? After all a 30 year IRU goes on the balance sheet as an asset and you’re in IT not accounting… and certainly not legal.
Some data centers don’t have many options and there are Tiers that regulate diversity in power, bandwidth, cooling, etc. Geography and Right of Ways can dictate a data center’s ability to be diverse. Data Center operators do their best to get the right pricing but a mistake can be crucial, locking them into a price for a long term… Long enough to no longer be competitive and shift with the market’s rates.
It’s my opinion that carrier agnostic data center operators provide the most value to the client. In addition, carriers that build out on spec rather than waiting for orders are typically more progressive and have tools to give them an edge on their opponents who are also selling bandwidth.
Today’s carriers are going as low as twenty-five cents per meg in data centers although some operators are still locked in at rates between $3-7 per meg wholesale. Those rates make solutions like AWS make more sense even though your bandwidth costs in an AWS environment can fluctuate from month to month. When other soft costs like travel to and from the data center and downtime associated with swapping out parts, running cross connects, etc., are factored in, that high bandwidth cost can really alter a company’s strategy.
It’s always helpful after getting your bandwidth turned up, to do a trace route to determine the path redundancy and ensure you get what you ordered. It’s relatively easy to cancel a newly installed circuit in a data center but that’s for a customer. The data center itself will typically eat the cost of construction or guarantee a long enough term for the build out to make sense for the carrier. In my experience that always worked out best when said data center works with its anchor tenant to establish a mutually beneficial arrangements for the customer, carrier, and data center.