Prometheus is a helpful tool to get started with metrics monitoring. Prometheus allows you to ingest metrics for a few components of your stack easily. Unfortunately, there are limits with a simple Prometheus setup. As you grow and need scale, Prometheus doesn’t give you a single pane of glass for all your metrics, persistent data storage, robust data analysis tools, smart-flexible alerting, or an inherent high availability configuration. The Wavefront platform can quickly solve these problems, so you don’t have to spend your time and energy building new services to supplement Prometheus.
Understanding Prometheus
Prometheus is an open source software project that scrapes HTTP endpoints to collect metrics. Services like Kubernetes and Docker have these metrics emitting HTTP endpoints built in. Other services and applications need to be instrumented or attached to a Prometheus exporter to emit metrics. When you run a Prometheus server, it collects metrics from all of the HTTP endpoints you configure and stores them locally. Your Prometheus server also provides a simple web UI where you can graph any of the metrics you collect. Alerts can be managed via a separate service but must be created in a configuration file. If you want dashboards, you’ll need to use a different tool. This setup can work for a very small system, but soon you’ll need more.
Wavefront’s Prometheus Integrations
Wavefront treats Prometheus metrics like any other data. This means that both of our Prometheus integrations offer access to our reliable database that supports fast querying, analytics-driven dashboards and alerting interface. These Wavefront features will dramatically upgrade any Prometheus setup.
Our Prometheus integrations have different use cases. The first integration is excellent for monitoring static environments where the list of HTTP endpoints doesn’t need to change often. The second integration is perfect for dynamic environments in which the collection of services that are running is continuously changing.
Integration #1: Static
The static integration completely replaces your Prometheus server by scraping the HTTP endpoints directly, converting the data to the Wavefront format, and sending it into Wavefront. This integration can be set up in three easy steps.
First: Use our install script to configure a Telegraf agent. This will replace your Prometheus server as the HTTP endpoint scraper. (If you don’t have a Wavefront proxy installed you can do that here too.)
Second: Configure our Prometheus plugin to scrape the URLs that you choose.
Third: Restart Telegraf.
Integration #2: Dynamic
Our second integration allows you to keep your Prometheus instances running, preserving your existing service discovery mechanism. With this integration Prometheus acts as a collector agent, forwarding metrics into Wavefront so they can be seen in the global context of your system. To use this integration all you’ll need to set up is the Wavefront Prometheus storage adapter.
First: Make sure you have a Wavefront proxy installed to forward metrics from Prometheus to Wavefront. Check our documentation for more information on our proxy.
Second: Run the Wavefront Prometheus storage adapter. This can be done either by downloading and building the source or by using the docker image we provide.
Third: Configure Prometheus to talk to the Wavefront Prometheus storage adapter. This is a one line change in each of your prometheus.yml files. You’ll need to restart each Prometheus instance after this change.
Summary
Wavefront’s Prometheus integrations give you everything you need to quickly upgrade your monitoring with Wavefront. Once you have your Prometheus data in Wavefront, you’ll be able to take advantage of our intelligent dashboards, powerful querying, and smart-flexible alerting. Most importantly, Wavefront scales with your growing monitoring needs. Whether it’s a few Prometheus endpoints or hundreds of Prometheus servers, Wavefront is the clear way to upgrade your metrics monitoring.
Want to try Wavefront with your Prometheus data – check out our free trial.
Get Started with Wavefront Follow @FrankHattler Follow @WavefrontHQ
The post Integrating Prometheus with Wavefront for Easy Scaling and Failover appeared first on Wavefront by VMware.
