In my first blog in this series, I highlighted some of the main challenges teams face with trying to scale mainframe DevOps. To get past these hurdles, the key is to develop an incremental approach that enables teams to capture value along each step of the journey ...
It's not an overstatement to say that application performance can have a very real and significant impact on business performance. Employees expect and count on immediate access to applications to get work done — no matter what time it is, where they're located or what device they're using. And it's no secret that customers are perfectly willing to go to elsewhere if their expectations to engage through the platform of their choice and receive an optimized experience are not met.
To ensure that applications meet or exceed these high expectations, underlying IT infrastructures need to deliver continuously optimized performance. However, while delivering this optimized performance is critical, it is also growing increasingly challenging as infrastructures continue to get more dynamic, hybrid and complex.
To optimize application performance, IT teams need to proactively ensure the optimized performance of every underlying infrastructure component, including physical and virtual servers, networks, storage devices, databases, end-user services and cloud and big data environments. It is imperative that they not only identify and resolve issues quickly, but also pre-empt potential issue before there's an adverse impact on the user experience. Below are six key requirements that can help you achieve these objectives.
1. Capture and Report on Performance Metrics
Given the criticality of performance, simply reacting to alarms when infrastructure elements go down isn't sufficient. IT personnel need to be able to identify potential performance issues, and address them proactively.
2. Unify Views and Tools to Increase Visibility and Speed Issue Resolution
In many organizations, dozens of point monitoring tools have been implemented. As a result when issues arise, team members spend countless hours in triage meetings trying to figure out what went wrong and how to fix it, all while the user's experience suffers. IT operations teams need to adopt a single, unified view and architecture for monitoring their critical IT services, whether they're running in virtual, physical, cloud or big data environments.
3. Track User Experience
IT teams can be armed with vast amounts of performance metrics, but if they don't know what users are actually experiencing, they don't have the real performance picture. IT teams need to collect and analyze end to end infrastructure response times for real transactions, and also leverage capabilities for generating synthetic transactions that allow consistent tracking of transaction times, even during periods in which users aren't working with applications.
4. Employ Rigorous SLA Management
Once organizations have holistic monitoring in place, it is vital for IT teams to track performance and experience against service level agreements (SLAs). IT teams need to be able to track SLA compliance, immediately identify when potential issues arise and address them. Through tracking SLAs, IT organizations can measure their effectiveness in managing user experience and infrastructure performance. This measurement is also vital in accurately gauging team performance, setting goals and tracking progress.
5. Leverage Predictive and Actionable Analytics
To optimize both infrastructure performance as well as operational efficiency, IT teams should leverage predictive and actionable reporting capabilities. Actionable reports or alarms ensure your team is not spinning in circles chasing false alarms and predictive analytics provide early warnings of problems so that they can be corrected before end users are impacted.
6. Correlate IT and Non-IT Data to do Effective Capacity Planning
Addressing evolving customer expectations isn't just a matter of tracking IT data. By leveraging both IT and business data, IT teams can proactively identify bottlenecks and improve end-user experience. For example, correlating server CPU utilization metrics with even simple historical data, such as volumes of user logins or transactions, can provide meaningful insights into planning for capacity to accommodate future growth.
By addressing the above requirements, IT teams are better positioned to ensure critical business applications consistently deliver the phenomenal user experiences that will help their organizations succeed in the application economy.
Umair Khan is Product Marketing Manager for Infrastructure Management Solutions at CA Technologies.