When the performance of an application deteriorates, productivity is impacted and end-users will be left thoroughly frustrated.
In many cases, this may well be just the beginning of the problem: before the application can be fixed, the root of the issue must be identified and effectively communicated to all relevant parties.
Michael Allen, director of performance at Compuware, investigates how businesses can ensure issues are dealt with as painlessly as possible, rather than becoming lost in translation.
Anybody who uses a computer on a day-to-day basis is likely to tell you there`s no such thing as the perfect, infallible application. No matter how much a developer tests and polishes his creation before sending it out into the big wide world, there is no way to consider every possible occurrence encountered in the wider IT environment.
Once an application is deployed in its new technological home, it will undoubtedly be impacted by a range of constantly changing factors.
A combination of these may negatively impact the performance of an application, and the complex architecture of applications can make it very difficult to capture the right diagnostic information when problems do occur. End-users are left unable to complete their tasks, significantly reducing productivity. It is therefore key that the IT team is able to identify and address any problems as quickly as possible. Here in lies the true challenge.
Identifying the cause of the performance problem is often like finding a needle in a haystack. In research recently conducted by analyst house Forrester, almost a quarter of respondents indicated that resolving performance problems requires 10 or more staff. According to 85% of the sample, one of the primary reasons for involving so many people is difficulty in isolating the source of the problem.
In some cases the problem only occurs intermittently, and the technical support team may not even see it for themselves. They have to rely on end-user reports to narrow down the nature of the issue, making matters more confusing. The developer responsible for fixing the problem has no direct contact with the end-users at all and has to use the second-hand information provided by the support team.
This often represents a further stumbling block as the technical support team is often made up of individuals specialising in different areas, such as network management or databases, rather than working across all areas. At the first sign of trouble, the disparate units may start pointing the finger at each other and the resultant "blame game" just makes it harder to find the true cause of the problem and fix it as quickly as possible.
A united approach is clearly going to be more productive, though this requires a cultural shift within the IT department, and as such is no mean feat. The best way to affect this change would be to implement a solution that monitors the end-user experience, notifying the support team of any issues that arise, as they arise. This kind of system clearly indicates where the performance weakness is occurring so that there can be no argument or finger pointing. Such a change in approach, from reactive to proactive, offers a significant increase in the efficiency and speed of response to application issues.
The challenge now lies in promoting effective communication between the technical support team and the developer responsible for fixing the application weakness, often seemingly speaking entirely different languages. Fixing the error involves understanding its extent and cause. The developer will require a range of complex information in order to pinpoint the exact problem and understand how best to address it, and this may be beyond the support team`s understanding or experience with the technology.
Without accurate and comprehensive information, fixing the problem will be a much slower process than necessary, and may even involve completely shutting down the application, reducing business productivity. To avoid lengthy downtime and increase efficiency, businesses need to implement a system that analyses the performance of the application, capturing a range of diagnostic information and identifying the root of the problem, uniting developers and operations teams.
A solution that automatically understands the application`s complex coding and extracts the relevant data is vital. Some applications have been described as "like a black box" due to the lack of production visibility into the inner workings, and this kind of system effectively turns a light on, illuminating the problem and capturing the diagnostic information the developer needs, without any call for translation or complication. How best to address the performance problem can be determined by interpreting the results to locate the root of the issue. This is the most direct and cost-effective way to consistently identify and address performance problems.
Application performance issues are inevitable unfortunately. The crucial variable factor lies in how businesses respond to this threat. While encouraging a change in attitude is often an uphill battle, the business value of taking a proactive approach to troubleshooting is considerable. By streamlining and simplifying processes, application downtime and the associated costs will be reduced. Developers can get on with fixing the problem quickly, technical support staff can get on with more productive tasks than hunting for the elusive needle in a haystack, and end-users will see a faster return of functionality. Above all, the frustrations of the blame game and the language barrier will disappear.
Share