Introducing Metric Drains

Thomas Orozco on January 16, 2018

We are proud to announce the release of Enclave Metric Drains. Metric Drains are the metrics counterpart of Log Drains: you configure them as a destination, and Enclave will periodically publish metrics for your Containers to the Metric Drain.

As of today, supported metrics include CPU and RAM usage for all containers, and disk usage and I/O for databases. As for destinations, you can route metrics to InfluxDB (self-hosted on Enclave and third-party) and Datadog.

This feature greatly expands our previously-released Dashboard Container Metrics, and will be particularly useful for sophisticated use cases that require real-time or historical access to detailed metrics.

Indeed, unlike Dashboard Container Metrics, Metric Drains allow you to:

  • Review metrics across releases and as far back as you’d like: since the metrics are pushed to you, you are free to define your own retention policies.

  • Alert when metrics cross pre-defined thresholds of your choosing: here again, since we’re pushing metrics to you, you’re free to alert on them however you’d like (alerting is respectively available in Grafana and in Datadog).

  • Correlate metrics with other sources of information: at this time, Metric Drains support pushing metrics to InfluxDB as well as Datadog, where you might already be publishing other metrics (e.g. using an application performance monitoring tool).

To provision a new Metric Drain, navigate to Environment of your choice in the Dashboard, and open the Metric Drains tab.

Create Metric Drain

PS: To make it easier to get started with Metric Drains, we also added support for InfluxDB as a Database on Enclave. This lets you easily route metrics to a self-hosted InfluxDB database. We also have detailed instructions on deploying Grafana on Enclave to create beautiful Dashboards and set up monitoring using these metrics.