As it becomes clear that modern applications based on microservices are going to be a lot more complex to manage, there is a growing focus on observability. As a concept, observability goes beyond simple monitoring in that platforms that are employed to manage application performance should provide greater context. Today, every time there is an issue, a war room meeting of developers and IT operating teams will spend hours trying to locate the root cause of an issue. The insult to that injury is usually the fact that it typically takes only a few minutes to fix a problem once it’s discovered.
The primary platform IT teams rely on to monitor and, at least theoretically, observe applications are application performance management (APM) tools that are typically accessed as software-as-a-service (SaaS) applications. The reason that more organizations don’t make greater use of APM platforms is two-fold. The first is they have been historically expensive to employ. IT teams often limit their usage to mission-critical applications to limit costs. Secondly, an APM platform requires developers to implement their applications by adding agent software to collect metrics. Each agent needs to be deployed and maintained like any other piece of software. There’s no getting around the need for instrumentation, but a debate about the best way of collecting relevant application metrics is starting to emerge.
Putting devs in charge
Lightrun has developed agent software that developers embed within their applications to capture live production data. APM platforms, in contrast, enable IT operations teams to capture historical metrics that can be analyzed to troubleshoot issues after the fact, says Lightrun CEO Ian Peleg. Lightrun alternatively provides developers with a debugging tool that is informed by the metrics collected by Lightrun. That approach enables developers to discover and troubleshoot issues both before and after an application is deployed in a production environment.
The company is making available a free edition of its debugging tool that is limited to a single developer as part of an effort to make developers more aware that observability can now be shifted left toward developers versus continuing to rely on an IT operations or DevOps team to surface application insights, said Peleg. The cost of setting up and maintaining an entire platform just to observe applications is still cost prohibitive, notes Peleg.
APM approaches also require developers to exit the environment they used to create the application to log onto a separate platform, adds Peleg. “It’s constant context switching,” says Peleg.
The Lightrun approach, in comparison, only collects data from an application when a developer actually requires it, adds Peleg. That approach greatly reduces operational overhead.
Integrated app management
As developers assume more responsibility for the management of applications after they are deployed, it’s imperative to provide them the tools they need. Forcing them to adopt platforms that were designed for IT operation staff is not likely to meet with much success. Developers require tools that fit naturally within workflows made up of command line interfaces (CLIs) and integrated development environments (IDEs).
Developers are not necessarily unappreciative of the tools any IT operations team prefers. It’s just that observability of an application requires instant real-time feedback within the environment used to build it to have any meaningful impact.