ambari-grafana

Use ambari-grafana to visualize metrics exposed via Ambari in Grafana. The following service metrics are supported as of now: HDFS, YARN, HBase, Storm, Kafka, Flume, Accumulo, and Ambari Metrics

NOTE: Development of Ambari Grafana plugin has moved to: https://github.com/apache/ambari/tree/trunk/ambari-metrics/ambari-metrics-grafana

This repository is no longer actively maintained.

This has been tested with Ambari 2.1.2/2.2.0 + HDP 2.3.

ambari-grafana is licensed under the Apache License, Version 2.0.

Install Grafana

You can install Grafana on any host. It does not need to be co-located with Ambari Server. The only requirement is that it has network access to Ambari Server.

Install on CentOS/Red Hat:

sudo yum install https://grafanarel.s3.amazonaws.com/builds/grafana-2.6.0-1.x86_64.rpm

Install on Ubuntu/Debian:

wget https://grafanarel.s3.amazonaws.com/builds/grafana_2.6.0_amd64.deb
sudo apt-get install -y adduser libfontconfig
sudo dpkg -i grafana_2.6.0_amd64.deb

Install on SUSE/SLES:

sudo rpm -i --nodeps grafana-2.6.0-1.x86_64.rpm

Deploy ambari-grafana

sudo wget https://github.com/u39kun/ambari-grafana/raw/master/dist/ambari-grafana.tgz
sudo tar zxvf ambari-grafana.tgz -C /usr/share/grafana/public/app/plugins/datasource

Start Grafana

sudo service grafana-server start

Dockerized Grafana with Ambari datasource

Easy deployement for Docker users:

# create /var/lib/grafana as persistent volume storage
docker run -d -v /var/lib/grafana --name grafana-xxl-storage busybox:latest

# start grafana-xxl
docker run \
  -d \
  -p 3000:3000 \
  --name grafana-xxl \
  --volumes-from grafana-xxl-storage \
  monitoringartist/grafana-xxl

Visit Grafana XXL project for more details.

Create Ambari Data Source in Grafana UI

Access Grafana Web UI at http://grafana-host:3000 and log in as admin / admin.

Click on Data Sources in the left nav and click on Add New in the top nav:

Name: Choose your own data source name
Default: Recommended to make this the default data source so when you create new graphs, you don't have to select this Ambari Data Source every time
Type: Ambari
Cluster: This has to exactly match the cluster name in Ambari
Stack: HDP (or other stack name that is installed on the cluster)
Version: 2.3 (or other stack version that is installed on the cluster)
Url: URL of the Ambari Server, including the port E.g., http://1.2.3.4:8080
Access: proxy
Basic Auth: Enable
User: admin (or the name of any Ambari user that has read access to the cluster)
Password: password of the Ambari user above