The hosting business is a really great place to be these days. It may morph rapidly, but some things ring consistently clear. The dedicated server is one of those things. In the brief 10 years or so of my Internet hosting career, the way dedicated servers are delivered to customers and the way they are managed has gone from prop-jet to auto-pilot.

I got started in the dedicated hosting business under Lance Crosby (our current CEO) in October of 2003. At that time we had less than 100 employees, and it may have been less than 50. “Auto-provisioning” consisted of Lance offering pizza and cash bonuses for each white-box PC that we’d ‘ghost’ with a boot floppy using a networked imaging server (in between our support tasks of answering calls and responding to tickets). We used a popular product made by Norton* in those days to deliver servers as quickly as possible to feed what seemed like an endless demand. As time has gone by, our systems have vastly improved, and true automation is the rule now; Manual intervention, the exception.

Today, SoftLayer has 600+ employees, 80,000+ dedicated servers, 26,000+ customers and is on the verge of launching our international presence. One of the biggest reasons SoftLayer has been so successful is because we offer customers maximum control.

When you need online computing power these days, you have hundreds of choices. Most of your options are still centered on the general idea of the dedicated server, but there are variations depending on what needs are being targeted. Physical dedicated servers are now complimented by Cloud Compute Instances and Virtualized Instances to provide a more flexible platform to tailor to specific use cases. Some providers do better than others at integrating those platforms, and when we began incoporating cloud and dedicated in an integrated environment, our goal was to enable customers to control all aspects of their environment via a single ‘pane of glass,’ our customer portal.

If you’ve heard us talk about the features and functionality in the customer portal but have never seen how easy it is to actually navigate the interface, today’s your lucky day:

In a nutshell, you get the kind of server control that used to require driving down to the data center, popping on your parka and performing some troubleshooting in the freezing cold cage. You may have been troubleshooting hardware cooling, wiring or other hardware issues, and you’d usually need direct console access to all the different types of servers and devices loaded on your rack.

Thankfully, those days are gone.

Now you can order a dedicated server and have it online in 2-4 hours (or a Cloud Computing Instance which can be online in 15 minutes). You can configure their private network so that they can talk to each other seamlessly; you can add firewalls, load balancing, backup services, monitoring instantly. For maintenance issues, you have the convenience of BIOS-level access via the standard KVM over IP card included in every server so you can see low-level hardware indicators like fan speeds and core temperatures and perform soft IPMI reboots. Firmware upgrades for your hard drive, motherboard, or RAID card that once required the ever-hated floppy disk can now be done with a few button clicks, and speaking of RAID cards, our systems will report back on any change to an ideal status for your disk subsystem. If that weren’t enough, you’ve got monitoring alerts and bandwidth graphs to give you plenty of easy to reference eye-candy.

No more messy wiring, no more beeping UPS units, no more driving, no more parkas.

-Chris

*As a rather humorous aside: My former manager, Tim, got a call one night from one of the newer NOC staff. He was a systems guy, many of the internal systems were under his SysAdmin wing. He was awakened by a tech with broken English who informed him that his name was on the escalation procedures to be called whenever this server went down:

Laughter ensued and this story was told many times over beers at the High Tech pub.