When I was chosen by Lance to manage a "special project," I knew I had my work cut out for me. My mission is to redefine how data centers are built, so before I get too far in the creation of my data center facility, I thought it would be a good idea to get a quick refresher about how SoftLayer builds and runs its data centers:
You can disregard the references to Lance "requiring" me to go on the tour ... and the formality of Jon having to sign a note that said I successfully completed the tour. The references to those were just for dramatic effect since Lance and I are pretty much eye-to-eye about everything that needs to happen for PHIL's DC to succeed, and we both thought it would be good to have signed documentation that I went on a tour.
As I mentioned at the end of the tour, I didn't find the tour to be a complete waste of time because I was able to observe some of the biggest hurdles in building and maintaining a data center. Is redundancy really necessary ... or is it just redundant? What the heck did Jon say in the UPS room? How fast does a person have to run on a single treadmill for that treadmill to power 40,000 servers?
While I try to source my own "PHIL's DC" rack identifiers, my adoring public can take time to go through this video with a fine-toothed comb to suggest any potential data center innovations that you think I may have overlooked. I already have the most innovative and efficient data center designed in my head, but I'll consider crediting you if you share an idea ... even if (and by if, I mean "when") I've already thought of it.