Data Collection

Plug & Play

Already a Heroku user? Get started.

Visibility into Heroku Platform, Postgres and App runtime like never before.

Librato seamlessly integrates with Heroku's rich logging capabilities to access native router, application & database performance metrics.

Simple Plug and Play Install:

Provisioning the add-on

$ heroku addons:add librato

Enable per-dyno runtime metrics

$ heroku labs:enable log-runtime-metrics
Enabling log-runtime-metrics for ... done
$ heroku restart

Any language, any buildpack

No agent or language binding required. Log metrics work from any application that can print to the log stream.

Dashboards out of the box

Beautiful, interactive Postgres and Heroku app performance dashboards.

Measure anything, everything

Simply write your measurements to standard out (STDOUT) and we'll take care of the rest.

Insights without impact

No agent. No impact. Measure any part of your application with nothing more than a print statement.

Router

Router metrics provide details about the number of bytes sent and the amount of time spent connecting to and servicing each request. Librato calculates percentiles for each of these metrics that make it easy to understand and optimize the performance of your application.

Postgres

Postgres metrics provisioned by Heroku are designed to keep you informed about the usage and performance of your database. They are reported directly into a beautiful easy to understand dashboard.

Per-dyno Runtime

Per-dyno runtime metrics Dyno metrics keep you informed about how much memory and CPU are being used on each of your dynos. You can tell at a glance if the number or type of dynos needs to change, and track their behavior over time for capacity planning purposes.

Custom Metrics

Perhaps even more exciting, you can use the same log integration to submit any custom metrics through a simple, intuitive, logging format. If your application can write to standard out (stdout), than you can report custom metrics and have access to powerful statistical aggregations like distributed counting and percentiles.