Is the ELK Stack Still a Good Option for Logging and Monitoring?

The ELK stack is still quite popular, especially among people using Kubernetes. However, many Kubernetes users lean towards cloud providers that offer their own logging solutions, like AWS or GCP, which integrate with their logging services. If you want to enhance your metrics and distributed tracing, tools like OpenTelemetry are gaining traction and can send data to multiple backends, including Elasticsearch.

ELK is still relevant, but there are important caveats. A few years ago, Elasticsearch changed its licensing model, which upset a lot of users. This led companies to explore other options better suited for containerized environments, like the Grafana stack or managed solutions such as AWS OpenSearch or Splunk, especially since managing Elasticsearch can be a bit complex at scale.

While some companies still use the ELK stack effectively, especially for log analytics, many organizations have shifted to simpler or more cost-effective solutions for log observation. If you don't need the analytics that ELK offers, it might be worth exploring cheaper alternatives that can just handle logging or alerts.

I've noticed that many teams aren't adopting the ELK stack anymore. For example, my team switched to using OpenTelemetry and a self-hosted Grafana stack, which has turned out to be much more cost-effective compared to ELK's licensing and hosting fees. It's working great for us, especially for multi-tenancy options.