I reviewed a research report recently, "The State of Application Performance Management: Common Challenges, Diverse Solutions," by UBM TechWeb. It examined how certain IT trends, including virtualization and cloud computing, have increased application delivery complexity and management challenges. Besides containing a few interesting stats, it was also the basis for a new webcast hosted by InformationWeek, which I recommend to anyone seriously interested in this field.
According to the survey, most organizations do not have a single department that owns end-to-end application performance management. No surprise. Responsibility is typically spread out over multiple departments based on functional IT silos, or, given to one silo. It is difficult for one silo, often represented by a networking or systems team, to have the breadth of expertise required to prevent and solve problems given the complexity of the applications they are expected to support.The components of a single application may have different authors, and run on different physical and virtual systems that communicate across networks, which in turn may include internal and external segments with limited visibility. This is why otherwise sophisticated IT teams often take days or weeks (or more) to fix problems with key applications that affect their employees and clients. Successful APM solutions need to understand both the end user experience and the relevant infrastructure components.
I thought it was also interesting that the #1 contributor to performance problems was cited as being directly related to "reduced IT budgets" (which most can empathize with).
Strong APM solutions address this budget constraint issue by automating the tedious tasks of gathering and analyzing the large, disparate datasets. A small team that is armed appropriately can therefore support a more sophisticated set of applications without proportionately increasing its budget. But what do people do to improve performance once they acknowledge a problem? According to the survey, there is no clear pattern. Specific problems require specific investments. Hardware. Application development. Bandwidth. I'm sure that longer surveys would have a longer list.
It would be interesting to find out how many of these investments actually solved the problems they were meant to solve. But setting that question aside for a moment, it is clear from the survey and common sense that there is no one silver bullet. With so many choices of where to invest and the cost of prolonged performance issues, precise root cause isolation is needed as quickly as possible after problems are detected. If you happen to be in the business of selling IT solutions and services, having the ability to show the impact of your proposal on your client's critical applications is a big plus for the same reason.
The InformationWeek webcast expands on this research report with a broader discussion of the technical and organizational challenges facing IT teams responsible for application performance. Paul Korzeniowski from InformationWeek moderates between the two main speakers: OPNET’s PJ Malloy, a Harvard-trained physics wiz who has been leading successful APM product teams for over 10 years, and IT management guru Bojan Simic from TRAC Research. The recorded session is punctuated by Q&A from attendees of the live event last month.
Enjoy.
Posts
No comments:
Post a Comment