Americas

  • United States
by Tim Wilson

Hurdles remain for electronic software distribution

Opinion
Jul 16, 20034 mins
Data Center

* Shortcomings of electronic software distribution tools, and a survey

It wasn’t long ago that the phrase “software distribution” meant lacing up your shoelaces, schlepping a disk to each end user’s machine, and doing the installation on the spot. Despite all of the recent advances in electronic software distribution (ESD) technology, it seems like just yesterday that this “sneakernet” was the primary means of getting users up and running on a new application.

In fact, in some companies, it WAS just yesterday.

With all of the new ESD technology on the market – including tools for software testing, packaging, configuration management, and end user deployment – many corporations are still struggling to develop a structured approach to software distribution. While sneakernet is seldom built into the long-term strategic IT plan, it frequently remains a fallback for ESD systems that can’t handle the load or don’t work the way they should.

There are many reasons for the shortfall of enterprise ESD processes and/or technology, but most fall into one of three categories:

* Inability to handle the load. Although most corporations are not growing rapidly, most software deployment efforts are. The recent wave of security threats has given rise to a whole category of patch management tools and processes. In addition, most software vendors are now issuing new releases in smaller, less comprehensive versions and modules, making the releases cheaper and easier to buy – but more numerous. Although IT staffs are not growing, the software deployment task is becoming a greater challenge than ever.

* High rates of deployment failure. While there are many ESD tools available to help automate the software distribution process, many of them do not have a strong track record for successful installation of applications on the first attempt. Many automatic installations fail due to system incompatibilities, conflicts between applications, or lack of adequate memory on the target machine. And because even prepackaged software is typically customized for a specific workgroup or department, the complexity of software distribution – and the likelihood of failure in automated processes – continues to escalate.

* Lack of integration with other software development and management processes. While software distribution once was seen as a discrete function, most organizations are now finding it critical for ESD tools to accept a “handoff” from development and/or packaging tools, while also being able to hand off data to configuration management or applications management tools. When software distribution is kept as a separate “silo” from other software management functions, deployment glitches often follow.

These problems may once have been seen as minor nuisances – or even as acceptable hurdles – in the scheme of IT management. With the downturn in the economy, however, such inefficiencies may have a very real impact on the bottom line. High failure rates in software deployment can lead to significant employee downtime and delays in implementing new business processes, not to mention significant loss of precious IT staff and resources. In today’s economic environment, too much reliance on sneakernet can mean loss of end user productivity and business revenue.

Is your organization facing any of these software distribution problems, or perhaps others not mentioned here? If these issues are familiar – or even if you feel you’ve conquered the ESD monster – we’d like to have your input. Over the next two weeks (starting July 15), Enterprise Management Associates will be conducting a brief survey on electronic software distribution processes, strategies and tools. The survey, which takes just a few minutes to complete, asks some questions about your priorities in ESD, and the obstacles that you face.

Please take a moment and answer the questions at:

https://www.enterprisemanagement.com/survey/swd

Your responses will be kept confidential (even anonymous, if you prefer), and a randomly selected respondent will win a prize from EMA. If enough readers complete the survey, we’ll share the results in a future installment of this column.