These days it's a lot easier to start a software company than it used to be. With the plethora of cloud services available at the click of a mouse companies can focus less on infrastructure logistics and more on their product. In this way, startups have it easy.
Before I continue I have a confession to make - Marketpath is not a startup.
Wow! That felt good! You don't know how long I've been holding on to that secret.
Ok, so maybe you read the about us page and you already knew that. We've been around since 2001.
We've claimed that we're like a startup but, in reality, that is far from the truth. We are established, healthy, and revenue positive. But along with being healthy we have our share of legacy systems with legacy problems. More specifically, we have a lot of old equipment running our infrastructure.
In 2001, we purchased our first server. It was a Dell Poweredge 2500. I remember, because I bought it and was excited about the possibilities for that powerful server. General websites and web applications were a piece of cake. So, we added an email server and database to it. Then, over the next several years, we bought another server and then another. Initially, we only managed the servers. But then it changed.
In 2004, we moved to nFrame, a data center in Carmel, IN. Since then, we've had to buy all the firewalls, routers, switches, backup devices, metered PDU's, and all the other fun that comes along with managing a SaaS infrastructure. And that's just the hardware. There was a slew of software for everything from email to backups to databases.
For a small resourceful staff this was one heck of a burden, especially when things went wrong. We had to worry about single points of failure in infrastructure, server hard drives, backplanes, power supplies, fans, memory, SCSI cards, NIC's, and many, many other intricate and complicated devices going kaput. And then there was the software maintenance.
Me and other staff have spent many long nights standing above the air conditioning panel right in front of our rack. Luckily, in all these years, our longest outage for any website was limited to about 4 hours. And that's because when the you-know-what hits the fan, we got very creative.
Startups don't have to worry about infrastructure - at least in the same sense that Marketpath and every other software company that's been around for more than five years has had to. Infrastructure is a huge resource suck. Proper planning, implementation, and maintenance should be full time jobs. When everything seemed to be humming along nicely, some small issue was festering behind the scenes getting ready to rear its ugly face.
Startups can focus on building software and leave the infrastructure to the experts. Now there's Amazon Web Services, Microsoft Azure, Rackspace, and many more infrastructure-as-a-service (IaaS) companies. These companies handle all the gory details and monitoring of core hardware devices. They provide virtual servers and other services that can be plugged into easily.
Now, instead of having to high-tail it over to the local data center, customers simply jump on their iPad, launch the web portal and restart a server or provision more. Only the exact infrastructure necessary is deployed. In the past, we had to buy more hardware than we actually needed so we could scale up to meet demand. There were intricate formulas (that I never used) to calculate capacity needed now vs. capacity needed down the road.
Startups have this provisioning capability at their fingertips and can dedicate their time to building their business and building great software.
Don't get me wrong. Established companies have this too and often implement new projects in virtual environments. It's the legacy software that is the slow moving snail of adoption.
At Marketpath, we began our virtualization quest last summer. We implemented our internal project management and support system which was a simple database and web application. Not too bad.
Since, we've been