Skip to content


Subversion checkout URL

You can clone with
Download ZIP
Convert a formatted log stream into metrics
Go Shell

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.


Convert a formatted log stream into metrics.


L2met receives HTTP requests that contain a body of rfc5424 formatted data. Commonly data is drained into l2met by logplex or log-shuttle.

Once data is delivered, l2met extracts and parses the individual log lines using the log conventions and then stores the data in redis so that outlets can read the data and build metrics. The librato_outlet is the most popular and will put all of your metrics into your librato account. See the setup section to get started.

Log Conventions

L2met uses convention over configuration to build metrics.

$stdout.puts("measure=db.latency val=20")

Metrics Produced:

  • db.latency.min
  • db.latency.median
  • db.latency.perc95
  • db.latency.perc99
  • db.latency.max
  • db.latency.mean
  • db.latency.last
  • db.latency.count
  • db.latency.sum


High resolution buckets

By default, l2met will hold measurements in 1 minute buckets. This means that data visible in librato and graphite have a granularity of 1 minute. However, It is possible to to achieve a greater level of resolution. For example, you can get 1 second level resolution by appending a query parameter to your drain url. Notice that the resolution is specified in milliseconds.


You can run l2met in a multi-tenant mode or a single-user mode. The multi-tenant mode enables multiple drains with unique librato accounts. The single-user mode exposes 1 drain and maps to 1 librato account. This setup guide assumes single-user mode.

Create Librato Account

Once you have created the account, visit your settings page to grab you username and token. Keep this page open as you will need this data later in the setup.

Create a Heroku app.

$ git clone git://
$ cd l2met
$ heroku create your-l2met --buildpack git://
$ git push heroku master

Add database services.

I prefer redisgreen, however there are cheaper alternatives. Whichever provider you choose, ensure that you have set REDIS_URL properly.

$ heroku addons:add redisgreen:basic
$ heroku config:set REDIS_URL=$(heroku config -s | grep "^REDISGREEN_URL" | sed 's/REDISGREEN_URL=//')

Update Heroku config.

$ heroku config:set APP_NAME=your-l2met LIBRATO_TOKEN=abc

Scale processes.

$ heroku scale web=1 librato_outlet=1

If you wish to run more than 1 outlet, you will need to adjust a config variable in addition to scaling the process. For example, if you are running 2 librato_outlets:

$ heroku config:set NUM_OUTLET_PARTITIONS=2
$ heroku scale librato_outlet=2

Add drain to your Heroku app(s)

Now that you have created an l2met app, you can drain logs from other heroku apps into l2met.

$ heroku drains:add https://l2met:`uuid` -a your-app-that-needs-l2met


Install log-shuttle.

$ curl -o log-shuttle ""
$ chmod +x log-shuttle

Send data to l2met.

$ export LOGPLEX_URL=
$ echo 'measure="hello-from-your-l2met"' | log-shuttle

Now you should be able to see hello-from-your-l2met in your librato metrics web ui.

Something went wrong with that request. Please try again.