End User Response Time
Network uptime doesn’t always mean high application performance. Network Engineers have access to more data points for monitoring their networks than ever before. They have the ability to analyze device health, utilization, packet discards, and traffic flows, collecting this data into network management tools. Yet, despite having this visibility, they struggle to stay ahead of slow applications. The key to resolving and staying ahead of these trends is leveraging the visibility provided through End User Response Time analysis.
Why is End User Response Time Monitoring and Measurement so Important?
While network engineers seem to have a plethora of tools and systems to monitor the health of the network, there remains a gap between application and network monitoring that is driving a need for a different approach: end user monitoring. In addition, there is increasing pressure to better support end user needs and increase end user experience satisfaction by implementing end user monitoring.
As applications grow in complexity and scale, this will only get worse. In fact, 75% of IT organizations1 suffer from degraded (slow) business applications, which include business critical systems.
Even with advanced monitoring tools, 70 % of the time2 Network Engineers learn about slow applications from negative responses about the end user experience. This means that by the time they get involved, the business has already been impacted. Workforce productivity has a direct impact on the business, so IT organizations need to focus on end user monitoring and how end users experience IT services.
The key to resolving and staying ahead of these trends is leveraging the visibility provided through End User Response Time analysis.
1 IDG Research Services
What is important for end-user response time?
Tools that show green in terms of network health and uptime are important. They give an overall health of the system and show when and where an outage occurred. However, even with a healthy network, applications can still lag and cause a negative end user experience.
These analysis tools do not give the detail needed to face and resolve the increasingly complex application problems of today. Network Engineers need access into data streams and synthetic user transactions that will show them what responsibility the client, network, and server/application have in the overall End User Response Time. Without this end user monitoring data, problems can linger for a long time as the finger pointing and guesswork troubleshooting begin. This is especially the case when troubleshooting problems with cloud-based applications, since engineers do not have full access to the systems supporting them.
IT departments underestimate how expensive application performance problems really are. These issues cost the business in terms of reduced employee productivity and customer confidence, and they typically take a long time to resolve due to their complexity. To face these problems, it is not enough for network engineers to analyze their infrastructure, look for errors, and then toss the ball to another IT department once the network is exonerated. They need to go deeper and utilize the details provided by monitoring EURT.
Best Practices to improve end-user response time
The end user experience is delivered by successful and high performance transactions to the application. If a slow or faulty transaction occurs, even in the past, engineers can pinpoint exactly what the problem domain was and focus on data that leads to a resolution, not just a problem escalation or handoff.
Network engineers own the highways and byways through which traffic flows, giving them unique access to the end user traffic that accesses application systems. They need end user monitoring systems that will collect these data streams and perform automated analysis on the End User Response Time, especially in data centers where there are thousands of transactions per minute streaming in. Packet level visibility allows them to monitor how users are interacting with the servers, and most importantly, quickly identify which component of the system is slow:
- Application delay
- Server delay
- Network delay
- Client delay
In a hybrid environment where some applications are hosted in the cloud, these metrics can be determined by synthetic tests from an analysis system. If a slowdown occurs, engineers can be notified of the time of the event, as well as the problem domain that contributed to the root cause. Rather than relying on users to gauge the health of the application, engineers can proactively monitor these systems to ensure they are running at a high level of performance for end users.
High quality end user experience is delivered by high performance applications, which in the end drive high production businesses.