Cloud · SRE
Kubernetes and observability
Stable, monitored Kubernetes clusters, with metrics, logs and traces to find problems before users do.
Kubernetes on EKS or OpenShift and observability with Datadog, Prometheus and Grafana, with SRE practices to keep systems up.
Goal
Have stable, fast applications, with useful alerts and enough information to understand every failure.
Challenges
Clusters that are hard to maintain, too many or too few alerts, failures noticed only when users complain and rising infrastructure costs.
What we do
- Kubernetes: Clusters on EKS or OpenShift, configuration, security and node groups suited to each workload.
- Observability: Metrics, logs and traces with Datadog, Prometheus, Thanos and Grafana.
- Reliability: Alerts, incident analysis and performance tuning, including Java applications.
Data and tools
Kubernetes, Amazon EKS, Red Hat OpenShift, Istio, Datadog, Prometheus, Thanos, Grafana and the ELK and EFK stacks.
What you get
Configured and documented clusters, working dashboards and alerts, and a guide for responding to incidents.
Example
At an AI company, the team rebuilt the infrastructure on OpenShift and EKS, with monitoring shared by several teams in Prometheus and Thanos and separate node groups, including spot instances, to lower costs.
Problems show up in the alerts before they show up in complaints.
Kubernetes and observability courses
Tailored courses for operations and development teams, using the company’s own clusters and tools.
- Kubernetes: Deployments, configuration, security and troubleshooting.
- Observability: Metrics, logs, traces and dashboards in Datadog or Grafana.
- Incident response: Alerts, root-cause analysis and continuous improvement.