Posted by: S R Balasubramanian
application architecture, applications maintenance, applications management
Managing the complete IT infrastructure of any organization is not easy. Many a time we receive frantic calls from users crying hoarse that they are unable to go through their transactional process, are unable to take out reports or that they find it difficult to send mails or to access information. This is a complex issue that needs tackling; the CIO goes about diagnosing the fault in order to isolate the problem and set it right.
It sometimes is a nightmare. The CIO is not sure whether the issue is on the client, the network or somewhere in the back-end infrastructure. He brings to bear his problem solving acumen and tries to go through the problem methodically trying to eliminate various issues so that he can zero in on the root cause. When he has a large and multi-layered IT infrastructure he has to drill down to each layer trying to discover the problem. This often takes too long to diagnose and resolve incidents. As users call him up frantically, he loses his peace and composure. When introducing new applications he has no means to ensure that the application performance will be as good as the developers promised and spends weeks after the roll-out to resolve issues.
Use of specific tools
It is not as if the CIO is devoid of resources. He would have bought performance measurement tools for database, applications or the network and use them to diagnose faults. When in problem he uses these tools to find out faults if any in each of these areas in order to identify and correct them. This takes time since he can’t visualize the problem in a holistic manner but has to wait to find out whether each of the areas are good or have an issue which needs tackling.
Applications have traditionally been delivered from data centers. Therefore traditional performance management has focused on data center infrastructure but delivery of today’s application deals with a complex data center architecture (including servers, storage, virtualization platforms, security layers etc.), multi-tier application architecture and an expanding mix of cloud services, third party services, mobile carriers, browsers etc. The CIO therefore needs a host of tools to manage and administer the complete infrastructure. He has tackle the issue on two fronts – to investigate whether the health of the infrastructure is alright or if not, where is the problem; and secondly, to run, diagnose, and fix problems vis-à-vis the application package if that is where the fault lies.
A holistic solution
It was therefore necessary to have a set of tools which could help IT administrators to proactively scan the environment to of any potential bottlenecks and to diagnose and isolate faults in the entire chain so that they can be attended speedily. So a host of vendors, including Compuware, Computer Associates, HP, Quest, and others, came out and have been offering solutions to make this possible.
Most of these tools offer an end-to-end monitoring and management of application performance or in other words give a global view of application performance. They traverse the entire IT infrastructure to identify any blockage or bottleneck and also do a component deep dive monitoring to bring out specific problem in the application context. They incorporate a role-based dashboard to quickly assess health of applications. Further, analytics provide insight into faults so that corrective actions can be put in place. It is possible to define SLAs and track service level compliance over time. They follow industry-defined standards and therefore if you already have any application specific tool, it is possible to integrate it with the APM suite.
Should CIOs consider APM tools?
Choosing any solution has to be in the context of problems being encountered. IT setups in small and medium-sized businesses are not complex enough to warrant an elaborate solution. However, larger organizations with complex work environment and those sporting multi-layered IT infrastructure may well consider better methods to deal with application performance issues. Businesses are no longer tolerant of IT bottlenecks and it enjoins upon us to provide an uninterrupted work environment. APM suites therefore make a lot of sense.