Businesses are increasingly turning to purpose-built, custom applications to manage mission-critical processes. In fact, the average enterprise-level organization now has more than 460 custom applications to manage, according to the Cloud Security Alliance.
Of course, there are many advantages to deploying custom applications, including ensuring adherence to precise specifications and allowing for greater control over processes that enable differentiation in the marketplace. But there are also risks, especially when outages directly impact the customer experience, reputation, and revenue opportunities.
What is custom application monitoring?
In particular, monitoring the performance of custom applications and troubleshooting issues is challenging for many IT operations teams. Some organizations may not even know that their applications are failing until they start to get notifications from impacted employees and customers – and by then, the damage is already done. So, what makes custom application monitoring so challenging? And how can organizations become more proactive in assuring their performance?
Let’s explore a few different types of custom applications, how application performance monitoring is traditionally done, and discuss the benefits of adopting a packet-based approach for large organizations with distributed infrastructure and diverse user locations.
Examples of custom applications and their uses across industries
When organizations are trying to solve a particular challenge – such as how to let customers file a request or receive a status update – off-the-shelf applications may or may not fully meet their needs. There may be unique challenges, depending on the industry, that require the build-out of custom applications. These can better address critical business-level requirements like regulatory compliance, connectivity to legacy systems, and operation across varying network environments.
Consider a few examples:
- The operations team within a manufacturing company wants to be automatically alerted when production issues arise at an aging, remote facility.
- An insurance company wants to roll out a mobile application to allow customers to update the status of claims, track payments, and reach out for 24/7 support.
- A hotel or airline brand needs to transform its loyalty rewards program and give customers the ability to sign up for special offers, check point balances, book reservations, and track award progress status.
In the cases above, off-the-shelf applications simply won’t do. However, monitoring application performance errors across dozens of shoppers in retail stores nationwide or the production status of a remote factory with lots of Industrial IoT devices. That's a task that's easier said than done.
Challenges with custom application monitoring
With so many custom applications to monitor in today’s vastly connected enterprises, often, the problems are as diverse as the solutions.
To draw more on the manufacturing example above, downtime lasting just a few minutes in one area of production can cause backups and costly delays downstream. However, application performance challenges could be tied to everything from the misconfiguration of fault-sensitive IoT devices to network issues, such as lag between the factory and cloud-based resources, throwing production out of sync. Troubleshooting either possible issue is difficult and time consuming, and it still may not discover the source of the problem, elongating resolution time as more investigation is required.
Additionally, IT operations teams are challenged by the reality that while they may not have been originally responsible for developing the custom applications used by their various lines of business, they share the responsibility of maintaining their ongoing availability and ensuring adherence to pre-set service level agreements. This task is incredibly challenging as custom applications are not well-known for having the best documentation available, especially when different developers have touched them frequently over the years.
Finally, the monitoring tool in use must have customization capabilities to configure the analysis of a custom application, which is often not the case. In cases where it is possible to configure monitoring custom applications, it may be time-consuming and delayed due to other IT priorities. When an organization has hundreds of custom applications deployed across its line of business, this can feel like a herculean task. Moreover, IT may not even learn of a custom application's existence until it begins experiencing issues, forcing them to play catchup when asked to help troubleshoot on top of their other duties.
For reference, here are just a few ways IT may become aware of applications:
- DevOps communicates when a new app is introduced
- A business unit notifies NetOps of a new custom app contract and collaborates on its introduction and support
- IT discovers the app by receiving a trouble report involving a custom application
- When troubleshooting something else, IT finds evidence of an unreported app
Adopting a packet-based approach to monitoring applications
With so many applications to monitor, many organizations are adopting an approach that uses deep packet inspection for application performance monitoring. This vendor-agnostic approach helps companies set baselines for application performance based on network traffic, create easy-to-read dashboards based on expected performance, and trigger-based alerts when application behavior veers outside of normal parameters. It also allows IT operations teams to drill into performance and identify the root cause of hard-to-diagnose problems, such as latency, intermittent bandwidth issues, broken dependencies due to migration of hosted resources, and more.
Why monitoring custom apps is critical
In summary, custom applications are becoming increasingly important to day-to-day business operations. Organizations across a range of industries are relying on their performance to manage mission-critical processes. However, the reality is that custom applications are only sometimes well-maintained or reported to IT, and they can be difficult to troubleshoot when they fail. When that task falls to IT, serious consequences are at stake from a customer experience, reputation, and revenue perspective.
Maintaining operational continuity is a matter of becoming more proactive in the monitoring and maintenance of custom applications. Doing this effectively requires greater visibility into both the applications that IT knows about already and the growing list of applications created by their lines of business in order to prevent downtime and performance issues. Only a robust packet-based approach – enabling comprehensive visibility across the enterprise and its applications – can truly meet this challenge today and in the future.
Eileen Haggerty is the area vice president of product and solutions marketing at NETSCOUT.