After speaking to thousands of APM users during my time with Gartner, I have seen 5 key issues that cause APM failures ...
More than half (63%) of senior IT operations executives are dissatisfied with their Application Performance Monitoring (APM) solutions, and 75% are dissatisfied with their Business Service Monitoring (BSM) solutions, according to a new BlueStripe survey of Fortune 500 companies.
While reasons vary, a common theme is the inability of these tools to keep pace with the make-up of applications both in the data center and within public and hybrid cloud environments.
Top reasons for dissatisfaction with APM tools, according to the survey, include an inability to support all applications or track all application components; metrics that are too developer-centric; difficult tool integration; and the simple fact that the tools do not actually help IT solve problems.
The problems cited with BSM tools include manpower requirements to keep service models up to date; lack of root cause analysis; too many alerts; difficult integration with other tools; and limited alerting for service level issues.
The survey highlighted three key trends in IT Operations:
- Current IT Operations processes for application monitoring and problem solving are both ineffective and manpower intensive
- IT Operations leaders are dissatisfied with their current set of performance monitoring and management tools
- Enterprise companies are hesitant to move mission-critical transactional applications to the cloud until processes and tools become more effective
“As companies continue to incorporate new technologies into their applications, the inability of conventional APM and BSM tools to keep up is taking its toll on IT Operations,” said Chris Neal, BlueStripe co-founder and CEO. “We were surprised to learn that in 2013, 81 percent of companies still have more than a quarter of their application issues go un-resolved, even with APM and BSM tools.”
Additional results from the survey:
- 68% of respondents reported failing to identify at least 1 in 10 business impacting incidents before users did
- 36% of respondents reported learning about more than 25% of problems from end user complaints
- Only 8% of respondents have a monitoring framework that both aggregates alerts and provides appropriate application and service level context for interpreting and acting on those alerts
- 92% of of respondents either have fragmented monitoring, using separate tools, or basic integrated monitoring, which does not correlate alerts to service level issues
- 52% of respondents reported that the standard process for fixing outages is a bridge call - which in large organizations can involve more than 50 individuals
- Companies using bridge calls as the primary approach reported the lowest success rates, with only 14% solving outages quickly
- Companies that used smaller teams for problem solving reported a greater success rate, with 29% able to solve outages quickly
Survey results also indicated a sharp contrast in attitudes regarding virtualization and private cloud versus public and hybrid cloud deployments for critical applications. In last year’s (January 2012) survey, IT Operations executives indicated that they viewed virtualization and private cloud as “just another technology” to be managed within their application architecture. The 2013 results build on this, showing widespread adoption of virtualization and private cloud.
In contrast, attitudes toward public and hybrid cloud among large company IT operations executives were distinctly skeptical. Despite the rapid growth of public cloud services like Amazon Web Services (AWS) and Microsoft Azure, large companies are explicitly avoiding critical application deployments using public and hybrid cloud, in part due to the limited ability of APM and BSM tools to monitor and manage new technologies.
About the Survey
BlueStripe Software surveyed senior IT Operations executives at 166 large US-based companies in early 2013.