Infrastructure 2.0 on Ulitzer

For IT managers and end-users alike, today’s cloud computing infrastructure seems almost too simple and uncomplicated – compared to the first generation of complex and resource-heavy network and systems management tools.

A Rich Man’s Club
First of all, systems management was a rich man’s club. The domain belonged to the big four: IBM, HP, CA (formerly Computer Associates) and BMC Software and their legendary products – OpenView, Tivoli, Unicenter and BMC Patrol. They invented and dominated the market.

Not such a terrible thing – at first – because they offered amazing tools. But they were very expensive, unbelievably complex, and it required people with PhDs to operate, install and maintain. Installation alone took months…even, in some cases, years!

But there were no viable alternatives. So, professional-grade tools were only for big companies that could afford paying $1 million or more (and up-front, too). Alternatives grew, but—even today—the big four still own 40% of the market.

The Challengers
But about 10 years ago, a quiet bunch of IT revolutionaries, tired of expensive, complicated operating systems and the resources needed to support them, created a new breed of network management software and network monitoring systems. Pioneers like Solarwinds, Groundwork, Zoho (Adventnet), Hyperic and others offered 80-90% of the services of the big four at only 20% of the cost. They brought network and systems monitoring to the rest of us.

These companies are faring well, as I write this. For example, Solarwinds, which recently issued an IPO, according to Gartner, has a market capitalization of $1.5 billion, on sales of $93 million.

This is feeding a frenzy of interest by investors and those looking to acquire profit makers – defying the current recession.

Consider recent transactions: Compuware acquired Gomez for $295 million, Hyperic was bought by VMWare and NetQoS was bought by CA for $200 million.

Software Bound
But the challengers and the legacy systems both have one thing in common: their systems are based on software – which requires permanent maintenance efforts. For example, you still have to maintain servers, install setup software, make updates, allocate storage and archive data, setup notification channels, make monitoring failure-proof, and provide remote access.

Consider this metaphor: it’s like being stuck in a train on rails. You can only move in the direction the engineer (in this case the software vendor) is going. And you’re subject to their issues, too, among them multiple code version, legacy codes, long release cycles.

Flying, Rather than Taking the Train
But the development and growth of cloud computing is revolutionizing network management and monitoring services – whether it’s networks, websites, servers, VoIP, etc. It’s like turning in your train ticket for a plane ride.

The cloud changes the whole paradigm of network management and monitoring because it offers:

• multi-tenancy benefits, e.g. a single code base used by multiple customers
• quick release cycles
• behavioral analytics
• customers stay in the loop via frequent communications regarding downtime instances and other stats.

With cloud-based monitoring, customers are happy because there’s quick self-service, sign-up, deployment that takes minutes rather than weeks or months, dependable and constant upgrades to the latest version of software, and there are no maintenance issues or tasks to perform.

Yet, the cloud is not completely surrounded by blue skies. Yes, there are many businesses that are concerned about transferring their internal financial data from their own servers to the cloud. And, there have been recent episodes of cloud outages of consumer applications, such as Google’s email and blog services.

But I ask you… if businesses are comfortable with moving their customer data to the cloud, why would they resist migrating their server’s CPU metrics to the cloud?

Here’s my advice: get off the rails and onto the cloud.

Read the original blog entry...