diff --git a/installing/installing_bare_metal/bare-metal-postinstallation-configuration.adoc b/installing/installing_bare_metal/bare-metal-postinstallation-configuration.adoc index c5696a2cbb..62f471bec3 100644 --- a/installing/installing_bare_metal/bare-metal-postinstallation-configuration.adoc +++ b/installing/installing_bare_metal/bare-metal-postinstallation-configuration.adoc @@ -64,6 +64,17 @@ include::modules/nw-osp-services-external-load-balancer.adoc[leveloffset=+1] // Configuring a user-managed load balancer include::modules/nw-osp-configuring-external-load-balancer.adoc[leveloffset=+2] +// Hardware metrics in the Monitoring stack +include::modules/bm-about-ipe.adoc[leveloffset=+1] + +// Adding hardware metrics to Prometheus observability stack +include::modules/bm-configuring-ipe.adoc[leveloffset=+2] + +[role="_additional-resources"] +.Additional resources + +* xref:../../nodes/clusters/nodes-cluster-enabling-features.adoc#nodes-cluster-enabling-features[Enabling features using feature gates] + // Using the Bare Metal Operator include::modules/bmo-config-using-bare-metal-operator.adoc[leveloffset=+1] diff --git a/modules/bm-about-ipe.adoc b/modules/bm-about-ipe.adoc new file mode 100644 index 0000000000..bd255c61b6 --- /dev/null +++ b/modules/bm-about-ipe.adoc @@ -0,0 +1,23 @@ +// This is included in the following assemblies: +// +// * installing/installing_bare_metal/bare-metal-postinstallation-configuration.adoc + +:_mod-docs-content-type: CONCEPT +[id="bm-about-ipe_{context}"] += Hardware metrics in the Monitoring stack + +[role="_abstract"] +Hardware metrics can be exported to the cluster by enabling the Ironic Prometheus Exporter (IPE). + +IPE is a tool that exposes the hardware sensor data of cluster nodes in the Prometheus format. +When you enable IPE in your cluster, the tool collects data from the baseboard management controller (BMC) of each node and exports the data to the cluster's monitoring stack. + +[NOTE] +==== +This method of collecting hardware metrics works only on Redfish-compatible BMCs. +==== + +You can then view these hardware metrics alongside other metrics in the **Observe** tab of the web console. + +:FeatureName: Monitoring bare metal hardware metrics +include::snippets/technology-preview.adoc[] \ No newline at end of file diff --git a/modules/bm-configuring-ipe.adoc b/modules/bm-configuring-ipe.adoc new file mode 100644 index 0000000000..139f358c08 --- /dev/null +++ b/modules/bm-configuring-ipe.adoc @@ -0,0 +1,84 @@ +// This is included in the following assemblies: +// +// * installing/installing_bare_metal/bare-metal-postinstallation-configuration.adoc + +:_mod-docs-content-type: PROCEDURE +[id="bm-configuring-ipe_{context}"] += Adding node hardware metrics to the Monitoring stack + +[role="_abstract"] +To access hardware metrics for your bare-metal nodes in the web console, enable the Ironic Prometheus Exporter in your cluster. + +.Prerequisites + +* You have enabled the `TechPreviewNoUpgrade` feature set in your cluster's `FeatureGate` custom resource (CR). +For more information, see "Enabling features using feature gates". + +* You bare-metal nodes use Redfish-compatible baseboard management controllers (BMCs). + +.Procedure + +. Enable the Ironic Prometheus Exporter by running the following command: ++ +[source,terminal] +---- +$ oc patch provisioning provisioning-configuration \ + --type=merge \ + -p '{"spec":{"prometheusExporter":{"enabled":true}}}' +---- + +. Optional: Configure the data collection interval by running the following command: ++ +[source,terminal] +---- +$ oc patch provisioning provisioning-configuration \ + --type=merge \ + -p '{"spec":{"prometheusExporter":{"sensorCollectionInterval":}}}' +---- ++ +Replace `` with the interval in seconds for collecting sensor data from BMCs. +The minimum value is `60`. + +. Optional: Disable default alerting rules for hardware metrics by running the following command: ++ +[source,terminal] +---- +$ oc patch provisioning provisioning-configuration \ + --type=merge \ + -p '{"spec":{"prometheusExporter":{"disableDefaultPrometheusRules":true}}}' +---- ++ +When `disableDefaultPrometheusRules` is set to `true`, the configuration prevents deployment of default alerting rules for hardware metrics. + +. Optional: Disable the Ironic Prometheus Exporter by running the following command: ++ +[source,terminal] +---- +$ oc patch provisioning provisioning-configuration \ + --type=merge \ + -p '{"spec":{"prometheusExporter":{"enabled":false}}}' +---- + +.Verification + +. From the web console, click *Observe* -> *Metrics* and enter "baremetal" into the **Expression** field. +Several autocomplete suggestions should appear, such as the following examples: ++ +`baremetal_power_status` ++ +`baremetal_temperature_status` ++ +`baremetal_drive_status` ++ +`baremetal_fan_status` + +. Select one of the autocomplete suggestions and click **Run Queries**. + +. Verify that the queried hardware metrics appear in the UI. + +. If you did not disable default alerting rules, view them by running the following command: ++ +[source,terminal] +---- +$ oc -n openshift-machine-api get promrule metal3-defaults -oyaml +---- \ No newline at end of file