Increasing Performance Visibility with EXFO-Brix Response Time Metrics
Users get frustrated when their favorite smart phone app fails to load or takes so long they forget which app they are waiting on. Does this look familiar?
As the importance of reliable network transmission increases every day, service providers and carriers must ensure that they meet customers' expectations. Any instance of customers experiencing degradation in service increase the risk of customer churn.
As carriers roll out more and more services on these networks, the need to monitor how users experience the network increases dramatically.
With that said, how do you ensure service availability to your subscribers?
Today, most carriers use “synthetic” tests to measure the actual performance of the network and services that ride on them. These performance metrics provide valuable information on how the network performs in delivering services. By measuring response time metrics, carriers can baseline and alert on conditions that fall outside the norm. With visibility of synthetic tests and the performance metrics available from network devices, carriers have two valuable data sets to help troubleshoot the source of service impacting problems. The end goal is to proactively detect degradations before they impact customers.
Carriers are able to layer service tests (think DNS, Ping, HTTP, FTP, etc…) on-top of the transport level providing a clear picture of where the issue resides. Let's use a DNS test as an example. A carrier may have a DNS and transport level test running from point A to point B, by being able to see the results of both tests in a single report, a carrier can see where the issue resides. If the time it took for the DNS test to resolve took an abnormally long time, but the transport level test time took the “normal” amount of time, it points to an issue with the DNS service. If however, the time it takes to resolve the DNS test goes up, as does the time for the transport test at the same time, this points the carrier to the issue being transport related, not DNS related.
Are synthetic tests alone enough?
Probe based solutions like the Brix Verifier probe from EXFO provide extremely valuable data and tremendous visibility for response time metrics. Unfortunately, EXFO BrixWorx and other probe report platforms do not provide a unified view of the response time metrics alongside the network device performance details. As a result, operations teams struggle with swivel chair management as they must look at multiple consoles to get a complete picture.
As a carrier you must prove to your customers that your network has the ability to deliver services and meet your customer's Service Level Agreements (SLAs), but do you lack the ability to measure how your network performs?
Carriers must unify these perspectives to give Operations teams one common integrated view of both data points (Probe data and Network data).
SevOne solves this problem by collecting bulk performance metrics from probe-based systems such as EXFO BrixWorx, Cisco, Alcatel-Lucent, and others and then providing a unified dashboard with all the metrics on one screen.
Once the data is in SevOne, carriers can report on the service delivery metrics alongside reports from the transport layer the service is riding on. With this single view of application delivery, you can baseline, threshold, and alert on the response time indicator to provide an early warning of performance problems.
Matt Goldberg is a SevOne Systems Engineer and (self-proclaimed) Marketeer.