We're excited to announce the first release of the Nucleus observability suite of tools. Our goal is to provide developers and ops teams with the tools they need to debug and monitor their infrastructure. We've taken care of installing and configuring both Prometheus and Grafana at the Environment and Service level so that teams can focus on building great products instead of installing and configuring infrastructure. We plan to continue to build out the native charts and graphs in the Nucleus dashboard while making it easy for teams to access Grafana where they can dig down into specific metrics.
Nucleus Environments are pre-configured Kubernetes clusters that Nucleus deploys and manages in your cloud account. You can create as many environments as you'd like with most customers typically creating dev, stage and prod. Currently, an environment is a single Kubernetes cluster (we have plans to let customers add multiple clusters to an environment in the future). When you create an environment you attach a node group (or multiple) that gives you the compute necessary to run your environment.
With Environment Metrics, teams can now quickly see Memory and CPU usage for all of the namespaces within the cluster. This is useful for several use-cases:
Performance Optimization - monitor the resource utilization of your cluster. This is helpful in identifying performance bottlenecks, detect resource-hungry applications, and optimize resource allocation.
Capacity Planning - understanding CPU and memory usage patterns helps teams in capacity planning and making informed decisions regarding scaling the cluster and adding or removing nodes.
Cost Optimization - in a cloud environment, CPU and memory usage directly impact cost. Monitoring these metrics enables you to identify over-provisioned or under-utilized resources, allowing you to optimize resource allocation and reduce unnecessary expenses.
Troubleshooting and Diagnostics - monitoring CPU and memory metrics can help teams identify applications or services that are using too many resources which can signal an issue and help teams troubleshoot.
As we continue to expand our node groups and compute offerings, metrics are going to become even more important in ensuring that you have the right resources for your services and environments.
Service Level Metrics
Developers deploy Nucleus Services into Nucleus Environments. Nucleus automates the auto-scaling of the services and spins up and spins down pods as needed. Similar to the Environment Metrics, Service Metrics also provide CPU and Memory metrics for each individual service pod. You can click on the pod in the graph to filter by that specific pod.
This is useful for many of the same reasons as in the environment level metrics. From managing costs to optimizing sizing and scaling rules, developers and teams can quickly understand how their services are performing and make any necessary updates. That can be updating Horizontal Scaling settings or creating a new Node Group with different Instance Types. It all starts with metrics.
Grafana is an rich data visualization layer for metrics collected from Prometheus and comes pre-built with a number of useful dashboards. We wanted to give developers and teams a first class, native experience in our dashboard with our Environment and Service Metrics and also give teams the ability to dig deep into Grafana. Now teams can instantly access Grafana right from the Nucleus dashboard through secure oAuth login.
Once you're in Grafana, you can customize dashboards, create panels and much more. You can even create custom dashboard and custom Prometheus queries. Whether you're an experienced Grafana user or brand new, there's a lot to explore and leverage in order to more confidently manage your infrastructure and services.
This is just the first release of many updates that we'll make to our Observability tools. Over the coming months, we're going to be adding more dashboards in order to help teams monitor and debug their infrastructure faster and using open source tools. From Service Dependency Graphs to Tracing, there's a lot more in the works. We'll continue to add to the native Nucleus experience and always keep the Grafana option there. If you're interested in trying out Nucleus, you can create a free account here . Stay tuned for more in the coming weeks!