Google Cloud Managed Service for Prometheus (2021)

What is Google Cloud Managed Service for Prometheus?

Kubernetes deployments play a crucial role in most enterprises’ IT services. As a result, enterprises tend to continuously deploy their applications on this platform. After an application has been deployed on Kubernetes, its key performance metrics need to be monitored, including user satisfaction, error rates, average response time, etc. Kubernetes in itself does not have application monitoring capabilities, but, you can leverage the capabilities of an open-source solution called Prometheus to achieve the same. Prometheus utilizes powerful tools such as Grafana to provide observability across Kubernetes deployments. As you scale up, it can become challenging to configure and manage Prometheus’ infrastructure and keep up with all the metrics from your successful services. This is where Google Cloud Managed Service for Prometheus comes in. 

Cloud Managed service for Prometheus is a fully managed storage and query service for Prometheus. It lets you scale your Prometheus-based monitoring of Kubernetes deployments at scale without having to manually manage it. It has been built on top of Monarch, which monitors all applications on Google. With this service, there is no need to use self-managed services such as Thanos

This service is fully managed by Google. This eliminates tedious and repetitive tasks such as:

  • Charting Prometheus servers,

  • Linking Prometheus servers to make a global data source, and

  • Managing complex scrape configurations.

It has been built on top of Monarch, a planet-scale monitoring system for all the applications on Google. This solution leverages the capabilities of this system to enable users to collect metrics at an incredible scale without any operational overhead. Since it shares a backend with the rest of Google Cloud’s monitoring well suited to complementarily work with Cloud Monitoring. For this reason, you can seamlessly collect metrics from all your services globally, i.e. global monitoring.

Managed Service for Prometheus – Possibilities

This service was designed as a casual replacement for Prometheus allowing new users to easily onboard. It is interoperable with Prometheus. This means that you can retain your existing Prometheus configs, Grafana dashboards, and PromQL-based rules and alerts. It ensures you maintain the portability of your workloads while moving to this service. 

In addition, this solution also provides managed collectors to simplify collecting metrics from your GKE clusters. That’s not all, it is not limited to only Kubernetes or GKE clusters. It has built-in support for multi-cloud and hybrid cloud, allowing users to monitor any environment in which Prometheus can run. 

The onboarding process is quite simple and can be completed in only two steps. First, you need to enable Managed Service for Prometheus on your GKE cluster then deploy Prometheus collectors. These collectors will be configured to write metrics directly to Managed Service for Prometheus. Next, you need to connect clients, for example, tools that you will use to visualize metrics from Prometheus. This can be Grafana or other tools you could use to read Prometheus metrics from Managed Service for Prometheus, and that’s it.

Previous
Previous

Google Distributed Cloud

Next
Next

Cloud Run for Anthos