Disclaimer: The following is not created, maintained, or endorsed by Docker. It is a reference guide to show a known and tested path for monitoring Docker Desktop in a production appropriate way
If you've setup all-in-one monitoring for test purposes and want to move on to a production ready architecture like below
then you can follow these steps with any prometheus remote write compatible software and a log collection endpoint or use the suggested configurate to match the all-in-one deployment.
Step 1: Install Loki via one of the following options to a dedicated location
Step 2: Install Mimir via one of the following options to a dedicated location
...
node_exporter:
enabled: true
relabel_configs:
- replacement: ${HOSTNAME}
target_label: instance
prometheus_remote_write:
- basic_auth:
password: "<Password>"
username: "<UserName>"
url: "http://SomeIPAddressOrDNSNameForMimir:9009/api/v1/push"
logs:
configs:
- clients:
- basic_auth:
password: "<Password>"
username: "<UserName>"
url: "http://SomeIpAddressOrDNSNameForLoki:3100/loki/api/v1/push"
...
...
metrics:
configs:
- name: integrations
remote_write:
- basic_auth:
password: "<Password>"
username: "<UserName>"
url: "http://SomeIPAddressOrDNSNameForMimir:9009/api/v1/push"
...
-
Refer to the Metrics Section Configuration Documentation for more metrics configuration options.
-
Refer to the Logs Section Configuration Documentation for more logs configuration options.
Step 4: Install Grafana to a dedicated location
