This is my most complex stack to date. It runs the unifi-poller (uPoller) application consisting of an influx database, grafana dashboard and the uPoller collector.
This has a lot of state (database, provisioning directory for granfa, state for grafana, config file for upoller, etc.)
This time i chose to store each of these in their own glusterfs volume mount using the gluster volumefs driver.
This has the advatages of creating folders that alread have the uid:gid of the grafan container (instead of messing around with chown and chmod)
The use of clustered volume allows swarm to start containers on any node as needed.
Once the containers have started for the first time you can provision grafan by copyying the content of the provisoning folder here into /gluster-vol1/gf_provoisioning and the upoller datasource and dashboards will auto provision. Make sure to edit influxdb.yml to update database password and username set in the stack. If it doesn't provision within a few minutes restart the container.
Only the granfan container needs to be published externally. It can be reached by any warm node IP or the keepalived IP. There is no need to conigure any form of name resolution for inter container communication.
Each container shares the same network and so each container can be used using the service name as this is provdided by docker. As such each containers can reach other by using servicename (as see in the - UP_INFLUXDB_URL=http://influxdb:8086 env var as an example,
(people over think DNS in containers! just remeber you can ping any service name from any container that shares a network!)
None, by default this template will result in a single replica of each container. It doesn't matter if containers start on different nodes.
version: "3"
services:
influxdb:
restart: always
image: influxdb:1.8
volumes:
- influxdb:/var/lib/influxdb
environment:
- INFLUXDB_DB=unifi
- INFLUXDB_ADMIN_USER=unifi
- INFLUXDB_ADMIN_PASSWORD=unifi
grafana:
image: grafana/grafana
restart: always
depends_on:
- influxdb
ports:
- '3000:3000'
volumes:
- grafana:/var/lib/grafana
- gf_provisioning:/etc/grafana/provisioning
environment:
- GF_SECURITY_ADMIN_USER=admin
- GF_INSTALL_PLUGINS=grafana-clock-panel,natel-discrete-panel,grafana-piechart-panel
- UP_INFLUXDB_URL=http://influxdb:8086
un-poller:
image: golift/unifi-poller:latest
restart: always
depends_on:
- influxdb
environment:
- UP_UNIFI_DEFAULT_URL=https://192.168.1.1 #IP of your UDM pro (if you have a USG please uPoller project docs)
- UP_UNIFI_DEFAULT_USER=<user you set in unifi>
- UP_UNIFI_DEFAULT_PASS=<password you set in unif>
- UP_UNIFI_DEFAULT_SAVE_SITES=true
- UP_UNIFI_DEFAULT_SITE_0=default
- UP_INFLUXDB_URL=http://influxdb:8086
volumes:
- poller:/config
volumes:
influxdb:
driver: gluster-vol1
grafana:
driver: gluster-vol1
poller:
driver: gluster-vol1
gf_provisioning:
driver: gluster-vol1
What do you intend to get out of this one ?