ent-dev-cloud-cover
Agile, Cloud

After 9 months, I delivered a new report

What started last fall as my biggest research project to date was successfully published last week on GigaOm Research. For non-subscribers, I’ve added the report to my work. I’d recommend you read the Executive Summary and if anything peaks your interest, you dive in deeper.

Titled Survey: Enterprise Development in the Cloud, the biggest different between this report and others is that first I had to design a survey to ask the right questions to inform the report. Red Hat’s hypothesis was that business and technical leaders didn’t realize the costs of waiting on infrastructure (and thus the value proposition of cloud to reduce wait time). So I attempted to quantify that.

The other unique thing about this survey is that it targeted people involved in “enterprise development”, all 408 respondents worked within IT of large and very large organizations (10,000+ people). To date it’s the only survey I’m aware that has this focus (please leave a comment if you know of others).

I worked with a few folks at Red Hat to get the 13 survey questions designed and then ResearchNow rolled it out and gathered the results in December. Then the real work begin sifting through the data looking for trends and patterns. Some were a little surprising, for example only 21% of respondents are developing apps in the cloud – a much lower figure than in other published reports (e.g. RightScales’ State of the Cloud survey has a lot higher adoption rate). Writing the initial draft and the back-and-forth of reviews and editing process took much of the winter and spring. But after 9 months, there’s finally something to show for all the effort!

For me one of the more interesting aspects of writing the report was organizing it around the key themes that emerged from the survey. ResearchNow provided a web-base way of slicing and dicing the survey results, so I could compare the project time to market for those users who use the cloud and those who don’t. I could cross-correlate data from different questions to see how they compared.

I won’t steal the thunder of the paper, but the evidence is pretty clear that cloud is improving time to market and productivity for enterprise development. What’s striking is the lower adoption rate today, but based on my recent experience this is rapidly changing as even the most conservative organizations are beginning to embrace the cloud, albeit slowly.

Give the paper a read and let me know what you think. Does this match your experiences? What are some other interesting observations you take away from this report?

Finally, I’ve got two more reports coming out later this summer and a speaking gig at Agile Richmond on July 15th. More on these to come…

Standard
Continuous Integration in the cloud
Agile, Cloud

continuous integration in the cloud

As part of my new gig as an analyst for the Virtualization Practice, I’m focused on the topic Agile Cloud Development. As discussed in my topic introduction, this covers the intersection of Agile Development and Cloud Computing. In determining how to approach this subject, I decided to break it down by types of products (aka tools) used in Agile Cloud Development.

First up in the Dev in the Cloud series is analysis on Continuous Integration in the Cloud. Overall I was happy with how this first article turned out. In writing it I spent quite a bit more time than estimated researching the various products and vendors in the space. What’s interesting to me, and didn’t make it into the article, is the two different types of cloud CI products in the marketplace.

The simplest are the CI solutions that work out of the box with minimal setup. Their primary selling point is simplicity of setup and ease of use. Most are tailored to a technology like Ruby or OSX and provide basic build and notification features. Examples include hostedci (OSX), tddium (Ruby) and Semaphore App (Ruby). These solutions are less expensive, less complex to integrate and can generally get teams up and running quickly. They are focused on developers and small start-ups.

At the other end of the spectrum are CI solutions that are a part of larger development Platform as a Service (dPaaS) offerings. The goal of dPaas products is to create completely integrated design-build-test-deploy solutions using their cloud software. In these products, CI is just one of the features along with features for testing, deployment, communication, management, issue tracking, collaboration, etc. Some are tailored to a technology, but most work with a variety of modern technologies including Java, Ruby, .NET and PHP. Examples include Bamboo, CloudBees (Java), CloudForge and Electric Commander. These solutions are more comprehensive and focused not just on developers but also managers and IT leaders who want a single solution from one vendor that integrates all the activities associated with software development.

These latter products are designed to help teams and organizations implement Continuous Delivery, a topic I’ll be tackling next.

If you’re currently using any of these tools (or others), please post a comment and let me know what you like or don’t like.

Standard