Kubernetes Application Performance Monitoring
Monitor your Kubernetes applications with zero setup or code changes. Get full tracing, RED Metrics and metrics with zero code changes.
Comprehensive Monitoring
Total coverage, no blindspots
- Every request
No sampling. 30x data compression means that you don't need to worry about storage.
- Every workload
Requests are collected at the linux kernel level. So you can see inside any container, regardless of whether you own it.
- Every cluster
Connect every cluster, use the same dashboards, compare between cluster to isolate issues
Up and running in minutes
Dashboards for every workload in less than 5 minutes
Metoro uses eBPF to generate all our data so you just need to install one Helm chart and the rest is done for you. Generated dashboards for every deployment, container and pod so you can get started immediately.
RED Metrics
Full RED metrics for every workload. Track number of requests, errors, latencies across your Kubernetes services. Identify performance bottlenecks and optimize response times.
Automatic Service Graph
See how your cluster works. Visualize your services and their dependencies, identify bottlenecks and optimize your service mesh.
Throughput Analysis
Analyze request rates and throughput patterns. Understand your Kubernetes application's usage and capacity needs.
Always-On Profiling
Metoro profiles all your containers so you can spot performance regressions down to the function level.
Cost Effective
Save on Your Observability Costs
Our efficient eBPF-based approach and optimized storage can cut your observability costs by up to 90%.
Example monthly cost comparison
Based on typical start-up usage
Hosts
Containers
Traces
* Approximate cost. Precise costs depend on the specific use case and any discounts.
** Assumes a single span is ~100 bytes. Updated 1st December 2024.
Technical Deep Dive
How does it work?
You install the Metoro helm chart into your cluster, this will create a daemonset which runs an agent pod on each node. The agent is responsible for collecting data from the nodes and sending it to the Metoro backend. The agent creates its telemetry data through eBPF programs that it injects into the kernel of the host it is running on.
What is an eBPF program?
eBPF is a way of running code inside the linux kernel in response to events that happen in the kernel. It's designed from the ground up to be as efficient and safe as possible. It's not possible for an eBPF program to crash the kernel. We instrument a number of events in the linux kernel, inspect them, process them and send them to the Metoro backend.
Why is eBPF useful?
- Doing telemetry at the kernel level means that we can collect data for every container without code changes. This is especially useful if you have a huge number of containers or if you run containers that you can't change. For example: user provided containers or third party containers.
- We can capture data without even restarting the application.
- The performance overhead is minimal in comparison to most user-space telemetry tools, and is typically less than 1% of CPU usage.
Ready to get started?
Start monitoring your Kubernetes applications in minutes with zero configuration required.