Pinpointing Application Performance Issues with Unified NPM
With employees and business partners working from home due to the pandemic, non-enterprise internet connections have exacerbated the issue of application performance. If services to your premises are degraded, you have recourse to your communication service providers. But if your users are dependent on mobile broadband, ADSL or NBN links, network performance is far from guaranteed—which can lead to frustration and reduced productivity.
However, application performance issues are not always caused by communication links. Sometimes issues and delays are created by the applications themselves, or the third-party services they depend on. In other instances, organisations delivering application services to anyone off their network have the need to identify and eliminate application performance issues for their own competitive advantage.
The challenge is how to comprehensively monitor, pinpoint and diagnose any issues across network links, the application and its dependencies, to avoid unproductive finger pointing and lengthy resolution times.
‘It’s a network problem!’
Time-sensitive applications—from ERP systems to online shopping to trading platforms—are dependent on optimum performance. It’s easy to simply attribute performance issues to the network or internet access, even in these days of cheaper bandwidth. In fact, degraded application performance is often due to other issues—so throwing more bandwidth at your WAN or cloud service links won’t necessarily fix it.
It is best practice to closely monitor and troubleshoot performance bottlenecks to isolate the true causes. Think about applications that rely on an external service as part of its operations. If your application goes off to another service such as check postcodes, ABNs, credit ratings or the like, you should have an SLA with that service provider. But if they are not meeting their SLAs, how can you quickly tell?
Breaking down a transaction, queries are sent to servers, which may query other servers, and back again. Then the response is provided to the user. If there is an unacceptable delay at any of these stages, it is critical to pinpoint the exact point so it can be resolved.
Success = technology, process, people
As in many complex situations, overcoming challenges efficiently and continuously depends on a combination of factors working in concert.
To gain visibility across performance bottlenecks, you need the right technology. Riverbed’s Unified Network Performance Management solution not only monitors network connections, it uses packet capture to be ‘application-aware’—providing you with a holistic view of the factors involved in application performance. When you can be accurate in your identification of issues, they’re faster to fix. Further, when you have evidence, you can ensure responsible third parties resolve them, without going through loops as fingers are pointed in all directions.
Stringent processes for monitoring, escalating, diagnosing and resolving issues—fast—are also essential. Over the years, Riverbed has developed the processes and methodologies for timely resolution and can share them with our customers.
Finally, your people need the right skills to go beyond network monitoring in these days of ‘working from anywhere’ to rapidly diagnose and resolve—or escalate to the right service provider—in real time. Again, Riverbed offers consulting services for skills transfer or ongoing support, as well as training services to get your team up to speed.
If you’d like to know more about the possibilities of application-aware network performance monitoring, our recent webinar Network Performance Metrics That Matter is available on demand and is a good start. Or, if you would like a demonstration of our Unified Network Performance Management solution, talk to your ICT service provider or contact us.