How to Better Manage Your Application Infrastructure

Recently, mValent conducted a survey of Fortune 1000 IT professionals to help understand issues and challenges associated with managing the complex, multilevel application infrastructure that supports mission-critical business applications and services to their internal and external customers. While conducted by mValent, this survey did not highlight any particular product or vendor. Rather, the survey focused on identifying and understanding key market trends.

Four key findings emerged from this survey, each of which has implications for IT managers as they seek to improve service levels. The four big-ticket items identified as priorities in the survey were the pressure to contain IT labor costs, the impact and cost of downtime and the associated drive to provide higher availability of applications, the imperative to provide better management over virtual server environments, and inefficiencies and delays introduced by configuration inconsistencies between environments (such as QA, staging, production, etc.).

Having the data is interesting, but what are the lessons learned? Be readily equipped to answer the boss’s question, “So, what are you going to do about it?” Here are some thoughts on actions you can take to make improvements against each of these findings.

Priority No. 1: Pressure to contain IT labor costs

Cost containment is the No. 1 business objective for executives in IT in 2008, possibly reflecting general concerns about the macro-level business climate. Forty-nine percent of respondents rated this as one of their top priorities, while another 23 percent tabbed “improving IT staff productivity” as a top priority. Cost and productivity are really two different ways to express the same concern, so both are included.

At this point, driving cost out of your IT environment (or improving your staff productivity) primarily means automating more processes to reduce the labor cost component of delivering and supporting applications. If you look across the various IT work processes, a few candidates for automation stand out. For example, the amount of time spent troubleshooting problems in preproduction or production is a significant time sink for many of the companies that participated in this survey.

So, too, is the time spent building out new environments to support new applications, new releases or expansions. Close to 80 percent of IT budgets are now spent on maintenance activities. Changing up this status quo can be an important step to draining inefficiencies out of the IT cost structure. There are many options to consider, whether you look at provisioning tools, configuration management products, process automation suites or others.

Priority No. 2: Impact and cost of downtime and the associated drive to provide higher availability of applications

About one-third of the respondents said that an hour of application downtime costs their company $100,000 or more. Ten percent said that the hourly cost for them was over a whopping $500,000. Clearly, there is a strong business case for any IT automation tool that reduces the incidence of application downtime. Since configuration inconsistencies are a significant source of application issues, you should consider targeting this area. Automated quality assurance and test tools would be an appropriate focal point to consider as well.

Priority No. 3: The imperative to provide better management over virtual server environments

Virtualization is one of the hottest trends in IT. Most companies have jumped on this bandwagon, but the survey results uncovered a critical issue deserving of focus for IT managers. While 37 percent said that implementing virtual servers made their environments less complex, the majority—63 percent—said it either made their environments more complex, made no difference at all, or they could not determine if virtualization made any difference. Clearly, this is a key issue for IT managers given all of the time and money thrown at virtualized environments.

Planning and performing server configuration management in the virtual server world will have to take into account new dynamics that did not previously require attention in the physical environment. Companies should purchase configuration management tools that can directly address the nuances introduced by a virtual server infrastructure.

Specifically, it would make sense for managers to add tools to their repertoire that allow them to manage an efficient set of virtual images, combined with a library of configuration settings tailored to various environments and applications.

Priority No. 4: Inefficiencies and delays introduced by configuration inconsistencies between environments (such as QA, staging, productions, etc.)

Our respondents cited the challenge of trying to troubleshoot configuration discrepancies as the No. 1 driver behind the investment in configuration management tools for the application infrastructure. Almost 40 percent said that was the case–and that’s surprising. The average Fortune 1000 IT organization supports literally hundreds of applications across a myriad of environments. Each environment, server instance and layer of software will have its own set of configuration parameters, typically without automated tools to manage them comprehensively. It follows naturally that these settings will quickly get out of sync and drift away from the company’s standards.

So, as you can see, there are major implications to improve service levels in application infrastructure management: Using some automated tool set for configuration, change and release management of these configuration settings makes good sense. These tools can either be commercially available or internally developed. But managing the often chaotic configuration environments will save valuable time for you and your staff. Time now spent on maintenance activities can be directed toward next frontier tasks, which can provide your company with efficiencies and competitive advantage.

 James Tauber is co-founder and chief technology officer at mValent. James has more than a decade of experience driving Web technologies and standards in both commercial and academic environments. He was one of the original invited experts who helped develop the XML standard in 1996. He is also a well-known open-source developer.

Prior to co-founding mValent, James was director of XML technology at Web services pioneer Bowstreet, where he was responsible for technical evangelism, standards and advanced technology. He served as Bowstreet's representative to the W3C and OASIS, where he served on numerous working groups and technical committees, and was chair of the OASIS Directory Services Markup Language TC.

James holds a bachelor of science degree in linguistics from the University of Western Australia. He is currently pursuing a Ph.D. He can be reached at jtauber@mvalent.com.