In the world of DevOps, developers ship code and operations deals with all the managing, monitoring and troubleshooting, but Google Cloud has found that it sometimes it makes more sense to put some of these responsibilities back in the hands of the developers. Today, the company announced the Beta release of Stackdriver Kubernetes Monitoring, a tool designed to give developers a comprehensive way of checking the health of their Kubernetes applications.
The company made the announcement at Kubecon, the conference related to all things Kubernetes going on this week in Copenhagen.
Arpana Sinha, who is the group product manager for Kubernetes at Google says that it’s about helping everyone in the monitoring chain, whether developers, operations or site reliability engineers (SREs). “What we are talking about at Kubecon is that there are a lot of challenges in monitoring modern cloud deployments, particularly hybrid deployments,” she said. This release brings together a couple of key pieces including Google Kubernetes Engine (GKE) and Prometheus, a popular open source tool for monitoring Kubernetes in on-prem installations.
What Google is doing with this new producing is bringing together logs, metrics and events across the various Kubernetes pieces including containers, pods, workloads and clusters, and combining that with signals from underlying infrastructure. That should result in giving developers a much more comprehensive view of their Kubernetes app — and that’s the goal with this release.
By connecting it to Prometheus, it’s really taking the cloud-native ethos to heart by giving a comprehensive view across on-prem and cloud-based Kubernetes installations. “Stackdriver is hosted solution from Google Cloud. It can monitor anything in Google Cloud, but Prometheus is tool you [would] use on prem. When you want single pane of glass, you can pull metrics from GCP and Prometheus on prem,” Sinha explained.
The release of this tool is another example of the company creating products based on ones they created in-house to monitor their own systems. Kubernetes itself was derived from an in-house container orchestration tool called Borg, the company had been using internally for years to manage its massive cloud applications like GMail and Google Docs.
This announcement comes on the heels of the recent Stackdrive APM announcement at the end of last month.