When we graduated Loki into a GA release last year, there were more than 137 contributors who already made more than 1,000 contributions to the project. We also added hosted Loki to the lineup of Grafana Cloud offerings after it proved to be stable internally for our ops cluster, storing 40TB and half a trillion log lines each month. There was, however, one persistent problem that kept surfacing, especially for developers who were writing applications in Go: The regex package was slow.

I’d like to share some of the best practices we’ve learned on our journey to battle performance issues with the Jaeger tracing tool. Some may say we are experts in logging. We log for a living, and have our log analytics service (which we based on open source ELK Stack) to prove it. We’ve mastered logging to the level where debugging and troubleshooting our system is a no-brainer.

Modern businesses rely on a complex ecosystem of digital networks, databases, applications, user devices, and more. The tools and platforms supporting this ecosystem undergo constant refinement and upgrades. However, the basic methods for monitoring infrastructure and application performance have remained unchanged, and logging is one such methods. Every computer, server, networking device, and application produces logs, a timestamped record of events within a system.

Modern businesses rely on a complex ecosystem of digital networks, databases, applications, user devices, and more. The tools and platforms supporting this ecosystem undergo constant refinement and upgrades. However, the basic methods for monitoring infrastructure and application performance have remained unchanged, and logging is one such methods. Every computer, server, networking device, and application produces logs, a timestamped record of events within a system.

The Stack Infra team at Elastic uses many services to provide downloads for projects like Elasticsearch and Kibana to our users. One of these services is Fastly, which helps us optimize delivery to regions around the world. Keeping an eye on the performance and behavior of our CDN is important to ensure we're operating at the level we expect.

Last week, the first OpenObservability conference took place. This event had amazing content contributions from open source project leaders, users, and influencers. We’ve seen massive growth and adoption in the open source observability space from the inspiring work being done across tracing, logging, and especially metrics. The new data stores and capabilities are growing at breakneck speed. There are more choices— yet more complexity—than ever before.