# Custom Metrics
---
# Two Primary Methods to get Metrics into Datadog From Your Applications
## DogStatsd
## Restful API
--
# DogStatsD
## Python implementation of statsD metric aggregation daemon from Etsy
## Enables Tags
## Sends metrics to the Forwarder
--
# API
## Sends metrics directly to app.datadoghq.com
--
# Which To Use?
## Using the API, every metric is a separate network transmission
* If there are many metrics being sent, this is a lot of overhead
## DogStatsD
* batches the metrics
* generates a histogram
* submits that on each flush of the forwarder
---
# Many libraries to choose from
##http://docs.datadoghq.com/libraries/
<!-- .element: style="background: none; box-shadow: none; width: 100%" -->
---
# Python
## https://github.com/DataDog/datadogpy
## http://datadogpy.readthedocs.io/en/latest/
--
# Using DogStatsD
from datadog import initialize, statsd
options = {
'statsd_host': <ipaddress of agent>,
'statsd_port': 8125
}
statsd.gauge(metricname, value, tags)
