GitLab Revamps Observability Strategy After Acquiring Opstrace

GitLab plans to make Observability a more integrated component of its Continuous Integration/Continuous Delivery (CI/CD) platform after acquiring Opstrace.

The open source project Opstrace reveals a horizontally scalable application programming interface (API) created by Cortex for Prometheus, an open source monitoring platform being developed under the auspices of the Cloud Native Computing Foundation (CNCF). It then uses the Loki API to collect log data.

DevOps teams can also point an existing instance of Prometheus to Opstrace or whatever instance of Fluentd or Promtail they use to collect log data, while supporting a variety of other APIs, such as the planned Datadog proxy. A command line interface (CLI) is provided to set the Opstrace instance that creates a console from which to manage log data collection.

Kenny Johnston, GitLab’s Senior Director of Product Management, said that as part of an ongoing effort to make monitoring more accessible, GitLab plans to integrate Opstrace virtually with GitLab Monitor for organizations using either the self-managed version of GitLab or a software-as-a-service (SaaS) platform.

Opstrace is currently collecting logs and metrics and is planning to support distributed traces. The goal is to extend the general scope of monitoring capabilities that GitLab already provides within its CI/CD platform, Johnston noted. He added that the goal is to make it simpler for DevOps teams to employ monitoring and data interrogation capabilities to improve workflow without having a separate monitoring platform.

GitLab is betting that as Prometheus continues to gain traction, more DevOps teams will prefer expanding their reach into the realm of observation rather than having to run a separate platform.

While observability has always been a core tenet of DevOps best practice, achieving it has proven to be a challenge. At best, most organizations can continuously monitor their IT environments based on pre-determined metrics that determine when a particular platform or application is performing within expectations. Monitoring combines metrics, logs, and traces – a specialized form of logging – to hardware applications in a way that makes it easy to solve problems without having to rely only on a limited set of pre-defined metrics created to monitor a particular process or function.

The rate at which DevOps teams embrace observability will naturally vary. In 2022, technology may not be the biggest hurdle so much as simply understanding the queries that can help DevOps teams better understand the root cause of an IT problem before a major disruption occurs. In the long term, machine learning algorithms are expected to take advantage of data collected by monitoring platforms to automatically identify problems that may lead to disruption long before it actually occurs.

One way or another, the ability to discern events even within the most complex IT environment is constantly improving. Whether this is accomplished as an extension of observational or by embracing true observational potential, it may not all be relevant – as long as the end goal is achieved.

Leave a Comment