Skip to content
This repository has been archived by the owner on Jan 13, 2022. It is now read-only.

Latest commit

 

History

History
63 lines (48 loc) · 1.67 KB

README.md

File metadata and controls

63 lines (48 loc) · 1.67 KB

Monitoring configuration

Requirements

This directory sets up a graphite server to track performance metrics.

Make sure you are running this in a firewalled environment.

We'll use docker to manage the installation, so follow the docker installation instructions:

https://docs.docker.com/engine/installation/linux/ubuntu/

Follow the instructions for the CE setup (Community Edition).

Next, install the hopsoft/graphite-statsd docker container, follow instructions at:

https://github.com/hopsoft/docker-graphite-statsd

Timeseries data can then be directed to UDP port 8125.

Mandatory config

The default config of the hopsoft/graphite-statsd docker container will likely cause your storage space to run out because of the amount of data being logged into statsd by the Django Workload. In order to solve this, please perform the steps below after starting the container. All commands should be run as root.

Obtain a shell in the container:

sudo docker exec -it graphite bash
cd opt/graphite/conf/

Add the following line to blacklist.conf:

^stats[^.]*\.benchmarkoutput\.

Edit the carbon.conf file to enable whitelisting. Search for the line containing USE_WHITELIST, uncomment it and set it to True:

USE_WHITELIST = True

Edit the retention policy in storage-schemas.conf:

[default_1min_for_1day]
pattern = .*
retentions = 10s:2h,1min:2d,10min:14d

Exit the docker container and restart it for the configurations to take effect:

exit
docker stop graphite
docker start graphite

Should the disk space fill up again, you can simply delete graphite’s database:

docker exec -it graphite bash
rm –rf /opt/graphite/storage/whisper/*