Our out-of-the-box solution allows you to monitor the health of your cluster with pre-built dashboards and easy-to-set alerts. It works with just a one-line install.
CPU and memory for pods/nodes, view limits, capacity, correlate to events, and set alerts on changes
Collect anything logged inside your cluster from internal applications and Kubernetes system components
View all incoming and outgoing HTTP requests alongside metadata. No APM needed.
K8s events dashboard, correlate to logs, alerting (ex crashloops, evictions, etc)
Monitor RPS, p95, and p99 latencies by microservices, including by URL path, alerts
Many engineering teams are working to improve monitoring and observability this year. While Kubernetes does have many benefits as a container orchestration system, implementing best practices can be tricky.
Here are five best practices that most teams should implement to improve visibility, cluster performance, and overall health:
Self-hosted open-source tools like Prometheus are often an easy starting point for engineering teams. But as you scale, self-hosted solutions often lag behind, break, and require increasing amounts of engineering resources to monitor the monitoring.
Consider using a managed solution like ContainIQ, or managed versions of popular open-source tools, like managed Prometheus. With a managed solution, teams are no longer responsible for self-hosting, updating, and maintaining their tooling.
Due in large part to improved kernel versions across major cloud providers, today’s engineering teams are able to monitor from the kernel and OS level, and not from the application level.
Using new technologies like eBPF, Kubernetes monitoring platforms like ContainIQ are able to deliver application-level insights from the kernel directly without the need to install application packages or middleware. By parsing the network packet from the socket directly, solutions like ContainIQ can deliver unique features, like latency by path/endpoint, instantly.
For example with ContainIQ, you can track how long your node.js application is taking to respond to HTTP requests from your users, ultimately allowing you to see which parts of your web application are slowest and to receive alerts when users are experiencing slowdowns.
We’ve all been there…scrolling through endless rows of logs looking to match up timestamps. Or even worse, losing cluster level logs at the absolutely worst time. Engineering teams should look for Kubernetes monitoring tools that show correlations between multiple metrics, events, and logs.
Platforms like ContainIQ are able to aggregate and efficiently store multiple data sources in one platform, allowing teams to correlate multiple disparate pieces of data together. For example, using ContainIQ, users are able to correlate events from the Kubernetes API to both application and cluster-level logs.
Alert fatigue is a real challenge for today’s engineering teams. While many SaaS and open source tools come default with alerting out of the box, not all alerts are created equal, and sorting through the noise wastes time and frustrates the team.
Teams should work to set intelligent alerts based on historical data and actions that have led to previous disruptions in end-user performance.
By using a SaaS solution like ContainIQ, users are able to view historical data and set more accurate alerts. ContainIQ’s Kubernetes monitoring allows users to deliver the alerts to a Slack channel.
More and more organizations are going multi-cloud. And while multi-cloud has many advantages, monitoring Kubernetes clusters across multiple cloud environments can be challenging.
Many open-source tools are not designed for multi-cluster support, and it becomes a tall task to manage multiple tools for separate cloud providers.
By using a platform like ContainIQ, users are able to aggregate multiple clusters across multiple cloud providers into one view. They can view, sort, and alert on this data from a single dashboard and alerting engine.
ContainIQ also allows users to filter based on specific clusters, namespaces, and more.