The monitoring system comprises of Prometheus and Grafana. Prometheus is configured as data collector and alert manager. Grafana is used for visualisation. Grafana can also be configured as alert manager. The alert manager is responsible for sending notifications via various communication channels like Email, Slack, PagerDuty and several others.
Prepare configuration repository¶
enable Prometheus/Grafana in
enable_grafana: "yes" enable_prometheus: "yes"
Data retention policies for Prometheus can be configured via command line arguments:
prometheus_cmdline_extras: "--storage.tsdb.retention.time=90d --storage.tsdb.retention.size=50GB"
Grafana unified alerting can be enabled as following. See Grafana documentation. With Grafana unified alerting the alerts from Prometheus can be handled within Grafana. Add the file
environments/kolla/files/overlays/grafana/grafana.iniwith the following content to the configuration repository.
[alerting] enabled = false [unified_alerting] enabled = true [smtp] enable = true from_address = email@example.com
Add Prometheus rule files and Grafana dashboards to the configuation repository. Clone the kolla-operations repository and copy all files from grafana and prometheus directory to the configuration repository. This will install a set of metrics definitions and alert rules for OpenStack services.
git clone https://github.com/osism/kolla-operations.git cp -r kolla-operations/grafana cfg-osism/environments/kolla/files/overlays/ cp -r kolla-operations/prometheus cfg-osism/environments/kolla/files/overlays/
If you want to use Prometheus alert manager, place the configuration in
environments/kolla/files/overlays/prometheus/prometheus-alertmanager.yml. See Prometheus alert manager documentation.
--- global: resolve_timeout: 5m smtp_require_tls: false smtp_smarthost: localhost:25 smtp_from: firstname.lastname@example.org route: receiver: default-receiver group_wait: 30s group_interval: 5m repeat_interval: 4h receivers: - name: default-receiver email_configs: - to: email@example.com send_resolved: true
Run Monitoring deployment¶
Run Prometheus and Grafana deployments:
osism-kolla deploy prometheus osism-kolla deploy grafana
The Grafana Dashboard will be available on the internal network at
The Prometheus web console will be available on the internal network at
The Prometheus Alert Manager web console will be available on the internal network at