Installs Grafana Cloud Agent on RedHat/CentOS or Debian/Ubuntu servers to collect observability data and sends it to Grafana Cloud.
This role installs and configures the latest version of Grafana Cloud Agent from GitHub releases. It also creates a Supervisor process.
It optionally installs Promtail, which is an agent which ships the contents of local logs to Grafana Cloud (Loki).
None. Other than an account on Grafana Cloud -> Create account.
Available variables are listed below, along with default values (see defaults/main.yml). You can get the values for your enviroment from the Grafana Cloud Portal.
prometheus_user: <username>
Each service in Grafana Cloud has a unique service id or user. Once in the Grafana Cloud Portal click on Prometheus to get the value you need to provide for prometheus_user.
grafana_api_key: <key>
You can generate a new API Key in the API Keys section of the Grafana Cloud Portal. The role has to be MetricsPublisher.
metric_label: <label>
target_host: <host>
target_port: <port>
Above variables define scraping targets and labels for your endpoints. Endpoint name is /metrics by default.
agent_version: latest
You can change the default version of the agent by specifying the agent_version variable. If not defined, will fallback to the latest Grafana Agent install. Change it to eg. "v0.18.1".
agent_location: /usr/local/bin
Location where the Grafana Cloud Agent's binary will be installed. The default location (/usr/local/bin) is preferred in systems where SELinux is enabled.
config_location: /etc/grafana
Location where the Grafana Cloud Agent's config will be stored. The default location (/etc/grafana) is preferred in systems where SELinux is enabled.
By default, this role will ensure Grafana Cloud Agent is running and enabled at boot.
loki_user: <username>
Each service in Grafana Cloud has a unique service id or user. Once in the Grafana Cloud Portal click on Loki to get the value you need to provide for loki_user. If this value is present, this role will install the Promtail agent and create a Supervisor process for it. It will scrape messages from /var/log and journald. For reference, see Journal Scraping.
None.
GPL-3.0 License
This role was created in 2021 by Nicolas Leiva.
In this fork we modified the role to be compatible for setting up the agent in containers.