Introducing Librato AWS CloudWatch (Beta) — Librato Blog

Introducing Librato AWS CloudWatch (Beta)


At the AWS Summit in San Francisco today, we’re introducing Librato for AWS, a turn-key solution for Amazon CloudWatch users who are looking for ways to get more out of the data CloudWatch is collecting for them.

The Beta version is available for AWS customers hosting applications in US-East and provides pre-configured instruments and dashboards for all of your EC2 metrics, with support for additional AWS regions and services under way.

Librato for AWS Benefits.

These are some of the benefits you’ll enjoy as compared to CloudWatch when using Librato for AWS:

  • One year data retention: We keep your data for a year, allowing you to do trend analysis, capacity planning and comparative analysis.
  • Persistent correlations: You can build instruments that show a number of metrics from multiple sources together. Instruments persist, so you have them available any time you need. This is a great tool for fire-fighting or just building a grouped set of metrics and saving them as an instrument that you can then display on a dashboard.
  • Configurable and shareable dashboards: You can build dashboards that combine instruments that you like to view together, for example showing all metrics for a server instance, a cluster, a database, or an application. Build as many as you want and easily share read-only versions with others.
  • Live update of instruments and dashboards: Instruments and dashboards are refreshed in real time, so you can easily monitor the health of your AWS resources on any display, from tablet to wall-mounted monitor.
  • Instrument and dashboard templates: If you’re provisioning large numbers of identical resources, such as server instances, there’s no need to build instruments and dashboards for each of them. You can build a curated set of instrument and dashboard templates for each set of resources of a specific type and dynamically instantiate one for a specific resource any time you need it.
  • Annotations: In addition to data streams you can send us annotation streams for aperiodic events, such as software deploys or configuration changes, that you want to correlate with metrics you are tracking. You can visualize annotations in the instruments of your choice.
  • Integration with popular collaboration tools: You can post snapshots of instruments in chat rooms like Campfire and Hipchat and send alerts to service integration tools like PagerDuty and OpsGenie.
  • Custom metrics with resolution up to 5 sec.: You can add custom metrics through one of our language bindings or one of the many open source monitoring tools can send data to Librato.


Our new standard pricing applies for data we collect from CloudWatch. This means that for 5 min. resolution data streams (the default resolution CloudWatch uses) you will be charged $0.05/mo. and for 1 min. resolution streams (used for detailed monitoring of EC2 instances) $0.10/mo.

Different from CloudWatch, we do not charge for Alerts or API calls and you can add as many custom metrics as you like, from $0.05/mo for 5 min. resolution to $0.30/mo for 5 sec. resolution data streams. This means that if you’re an extensive user of Alerts and Custom Metrics, you may even save money by using Librato for AWS!

How to get started

You can read the step by step instructions in the "Using Librato with Amazon CloudWatch" knowledge base article but in essence it's as simple as logging in to your Librato account (if you don't have one you can sign up for free), adding your read-only AWS credentials and picking a region and a service.

For this beta release we only support US-East and EC2 but we will add support for additional AWS regions and services in the near future. 

That’s all that’s needed to start seeing your CloudWatch metrics in the instruments and dashboards we’ve curated for each service and use all of the features our service provides for getting the most out of the data CloudWatch collects.

Let us know which regions / services you want us to integrate first by answering two questions in this survey: .