Here’s a common question we get from web developers these days: “are you getting rid of ColdFusion hosting?”

The answer to this, of course, is absolutely not.

We love ColdFusion and its developer community. And we’ve been supporting the platform since it was a product of Allaire (which was before it was a product of Macromedia), which is to say we’ve been supporting it for a long time.

With this said, however, we understand why our CF community would be worried.

To put it gently, supporting ColdFusion in a shared hosting environment is often difficult, and for some hosts (like GoDaddy), it’s no longer worth supporting at all.

But, unlike those hosts, we’re fully committed.

And to provide you with exhibit A, to prove to you that we’re not just feeding you some marketing mumbo-jumbo, we wanted to share with you the lengths we took to better support ColdFusion with it’s latest iteration: CF 11.

Some Context

The fundamental struggle with ColdFusion hosting in a shared environment is the issue with sporadic downtime, specifically from service restarts. This is an issue that all ColdFusion hosts deal with. It’s pretty much a nature of the platform in this type of environment.

As a result, we’ve put in a lot of work to stabilize the platform as much as possible, which we’ve been largely successful at doing. But we knew we needed to do more.

So, with CF 11, we went back to the drawing board and engineered an all-new configuration that addressed some of these trouble spots.

A New Way of Doing Things

To illustrate how our new environment handles requests, we can think about it as having two load balanced web servers. Until now, ColdFusion shared hosting has been delivered using a single ColdFusion instance, which serves all requests. Our new environment, on the other hand, provides redundancy using dual ColdFusion instances. The dual instances are load-balanced, and have session replication going between them. As a result, we’re now working with a high-availability, or fault-tolerant, ColdFusion environment.

So what does this mean for our customers? If we run into an issue with a ColdFusion instance, we’re able to gracefully restart it without having to bring down the entire service. In other words, when one instance becomes unavailable, all of the sessions and requests are shifted to the other instance.

Here’s an illustration of the new config:

Addressing Bad Neighbors

One of the biggest sore points for ColdFusion shared hosting is the issue of bad neighbors.

The fact is, problematic code from one app in a shared environment can impact the performance or availability of other sites or apps on the server. This is especially true in a ColdFusion environment.

Now, we have the ability to create small instances for specific customers while we troubleshoot their code-related issues. Think of this as a private ColdFusion instance that is only responsible for their code, thus removing their impact from the rest of the server.

The Newtek High Performance Cloud

On a final note, all of the benefits described to you above have the added benefit of living inside our high-performance cloud infrastructure, which removes any direct dependency of your service to specific hardware, among other benefits. If you’re not familiar with our cloud infrastructure, we invite you to read a little more about it in the following Microsoft Case Studies:

http://www.microsoft.com/casestudies/Case_Study_Detail.aspx?CaseStudyID=4000010251

http://www.microsoft.com/casestudies/Case_Study_Detail.aspx?CaseStudyID=4000010845

So what do you think of our new CF config? Let us know in the comments below.