A docker container for an complete graphite stack. Usable in combination with Grafana.

Supports python 3.

$ make
$ make run

[2018-09-10 13:47:20 +0000]  starting supervisor
[2018-09-10 13:50:12 +0000]  -----------------------------------------------------------
[2018-09-10 13:50:12 +0000]   graphite 1.1.4 (stable) / python 3.6.6 build: 2018-09-10
[2018-09-10 13:50:12 +0000]  -----------------------------------------------------------

Build

Your can use the included Makefile.

To build the Container: make build

To remove the builded Docker Image: make clean

Starts the Container: make run

Starts the Container with Login Shell: make shell

Entering the Container: make exec

Stop (but not kill): make stop

History make history

Docker Hub

You can find the Container also at DockerHub

get

docker pull bodsch/docker-graphite

run

docker run \
  --rm \
  --interactive \
  --tty \
  --publish=2003:2003 \
  --publish=7002:7002 \
  --publish=8088:8080 \
  --volume=/data/docker/graphite:/srv \
  --hostname=graphite \
  --name=graphite \
  bodsch/docker-graphite

Notes:

• Please make sure to specify a hostname, so that internal metrics of carbon are not saved with a temporary hostname

supported Environment Vars

• MEMCACHE_HOST
• MEMCACHE_PORT (default: 11211)
• USE_EXTERNAL_CARBON (default: false)
• GRAPHITE_SECRET_KEY

Ports

• 2003: the Carbon line receiver port (tcp and udp)
• 7002: the Carbon cache query port
• 8080: the Graphite-Web port

Limitations

The database store only dashboards and i think, grafana are the better tools for this part.

Conclusion, i do not use the database feature of graphite (sorry, guys). The sqlite database will only created in the /tmp directory and not used.

When your using a external carbon-writer (like go-carbon) you do not need the internal carbon. You can disable this part with USE_EXTERNAL_CARBON.

internatls

includes

• graphite-web
• whisper
• carbon-cache
• nginx

Storage Schemas and Retention Period

The configuration is located at /opt/graphite/conf/storage-schemas.conf and has this default entries:

[carbon]
pattern = ^carbon\.
retentions = 30s:7d,5m:30d,1h:720d

[default]
pattern = .*
retentions = 30s:6h,1m:15d,5m:30d,10m:240d

Each section has:

• a name, specified inside square brackets
• a regular expression, specified after pattern=
• a retention rate line, specified after retentions=

As example for retention we use the following

pattern = ^telegraf\.
retentions = 30s:7d,5m:30d,10m:1y

The regular expression pattern will match any metric that starts with telegraf.

Additionally, this example uses multiple retentions:

• each data point represents 30 seconds and we want to keep up to 7 days of data with such frequency
• all historical data for the last 30 days is stored in 5 minute intervals
• all historical data for the last year is stored in 10 minute intervals

To calculate the whisper file size, I can recommend this tool: whisper-calculator (gist)