IBM Cloud Docs
Monitoring for Satellite

Monitoring for Satellite

IBM Cloud Satellite® comes with basic tools to help you manage the health of your Satellite resources, such as reviewing location and host health. Additionally, you can integrate Satellite and other IBM Cloud resources with IBM Cloud® Monitoring to get a comprehensive view and tools to manage all your resources.

Monitoring for your Satellite location and for the IBM Cloud services that run in your location must be set up separately. For example, to collect metrics for your Satellite location setup, you enable a Monitoring instance to collect platform metrics in the same region that your location is managed from. Then, to collect metrics for a Red Hat OpenShift on IBM Cloud cluster that runs in your Satellite location, you create a monitoring agent in your cluster to automatically collect and forward pod metrics to a Monitoring instance. Note that you can use the same Monitoring instance to collect metrics for both your Satellite location and services that run in your Satellite location.

Understanding what is logged and monitored by default

By default, Satellite generates certain activity events and monitors the state of your location and host resources.

Auditing events for Satellite actions

See Auditing events for Satellite.

IBM monitoring to resolve and report location alerts

When you create a Satellite location and set up the location control plane, IBM automatically monitors and resolves certain alerts for issues with your location setup and host infrastructure. The following table describes different scenarios and the actions that IBM takes to address the scenarios.

Additionally, if you set up your Satellite location to forward logs to IBM Log Analysis, the messages and more detailed information from the IBM Monitoring component are captured and stored in your IBM Log Analysis instance.

IBM monitoring actions to address certain scenarios.
Scenario Action
Location control plane does not have a host in three separate zones. Check for attached, unassigned hosts in the location. If a host is available, assign the host to the location control plane for the missing zone, giving preference to hosts with a label that matches the zone.
Cluster capacity exceeds 80% in a zone. Prevent or allow Red Hat OpenShift clusters to be created. Assign available hosts to a location control plane for more compute resources.
Red Hat OpenShift clusters are in an unhealthy state. Resolve certain health issues with Red Hat OpenShift clusters.
Default monitoring tools like Prometheus do not work. Send alerts to your IBM Log Analysis instance and return a status message with further troubleshooting information.
Ingress subdomain registration fails. Alert IBM engineers to troubleshoot the issues further and return a status message with further troubleshooting information.

Viewing location, host, and cluster health

You can review the health of Satellite resources such as locations, hosts, clusters, and Kubernetes resources that run in clusters.

Viewing location health

When you set up a Satellite location, IBM Cloud monitors the host and reports back statuses that you can use to keep your location healthy. For more information, see IBM monitoring to resolve and report location alerts. For troubleshooting help, see Debugging location health.

You can review the host health from the Locations table in the Satellite console, or by running ibmcloud sat location ls.

Location health states
Health state Description
action required The location needs your attention. Check the status and message for more information, and try debugging your location.
completed Satellite completed the setup of the location control plane components on the hosts that you assigned to the control plane. The location is soon ready to be used.
completing Satellite is setting up the location control plane components on the hosts that you assigned to the control plane. Check back in a little while.
critical The Satellite location control plane needs your attention. Check the status and message for more information, and try debugging your location control plane.
failed Satellite did not successfully resolve issues in your location. For more information, see the status message.
host required The Satellite location is created, but you must assign hosts to the location control plane. Assign hosts in multiples of 3, such as 6, 9, or 12.
normal The Satellite location is ready to use.
provisioning The control plane for the Satellite is provisioning. You cannot assign hosts to other Satellite resources, such as clusters, in the location until the control plane is ready.
resolving Satellite is trying to resolve issues for you, such as by assigning available hosts to the control plane to relieve capacity issues. For more information, see the status message.

Viewing host health

When you attach hosts to a Satellite location, IBM Cloud monitors the host and reports back statuses that you can use to keep your hosts healthy. For more information, see IBM monitoring to resolve and report location alerts. For troubleshooting help, see Debugging host health.

You can review the host health from the Hosts table in the Satellite console, or by running ibmcloud sat host ls --location <location_name_or_ID>.

Host health states.
Health state Description
assigned The host is assigned to a Satellite resource, such as a location control plane or cluster. View the status for more information. If the status is -, the hosts did not complete the bootstrapping process to the Satellite resource. For hosts that you just assigned, wait an hour or so for the process to complete. If you still see the status, log in to the host to continue debugging.
health-pending The host is assigned and bootstrapped into the cluster as worker nodes that are provisioned and deployed. However, the health components that IBM sets up in the host cannot communicate status back to IBM Cloud. Make sure that your hosts meet the minimum host and network connectivity requirements and that the hosts are not blocked by a firewall in your infrastructure provider.
provisioning The host is attached to the Satellite location and is in the process of bootstrapping to become part of a Satellite resource, such as the worker node of a Red Hat OpenShift on IBM Cloud cluster. While the host reports a provisioning state, the worker node goes through the states of provisioning and deploying. You can log in to the host while in this state to view logs. See Logging in to a RHEL host machine to debug and Logging in to a RHCOS host machine to debug.
ready The host is attached to the Satellite location and ready to be assigned to a Satellite resource.
normal The host is assigned to a Satellite resource, such as a location control plane or cluster, and ready for usage.
reload-required The host is attached to the Satellite location, but requires a reload before it can be assigned to a Satellite resource. For example, you might have deleted a Satellite cluster, and now all the hosts from the cluster require a reload. To reload a host, you must remove the host from the location, reload the operating system in the underlying infrastructure provider, and attach the host back to the location.
unassigned The host is attached to the Satellite location and ready to be assigned to a Satellite resource. If you tried to assign the host unsuccessfully, see Cannot assign hosts to a cluster.
unknown The health of the host is unknown. If the host is unassigned, try assigning the host to a Satellite resource, such as a cluster. If the host is assigned, try debugging the host by following the steps in debugging the health of the host. If the host still has issues, try removing, updating, and reattaching the host.
unresponsive The host did not check in with the Satellite location control plane within the past 5 minutes. The host cannot be assigned when it is unresponsive. Try debugging the health of the host, particularly the network connectivity.

Viewing cluster health

To review the health of Red Hat OpenShift on IBM Cloud clusters that run in your Satellite location, see the Red Hat OpenShift on IBM Cloud documentation.

Viewing Kubernetes resources in clusters

When you add your clusters to Satellite Configuration, the Kubernetes resources are automatically added to an inventory that you can review. For more information, see Managing apps with Satellite configurations.

Adding clusters to Satellite Configuration does not automatically set up logging and monitoring solutions, such as IBM Log Analysis and IBM Cloud Monitoring.

Viewing Satellite Config registration status for clusters

You can view the registration status of clusters that are enabled for use with Satellite Config. Keep in mind that some clusters might be in a public cloud location, not your Satellite location.

  1. List clusters that are registered with Satellite Config. Note the output in the Status and Location columns.
    ibmcloud sat cluster ls
  2. Review the following Satellite Config registration statuses.
Host health states.
Status Description
active Satellite Config components for the location are installed in the cluster, and at least one resource is being watched.
inactive Satellite Config components were manually removed from the cluster, or are installed but are no longer responding to Satellite Config. For example, network connectivity might be disconnected. Existing resources, if any, continue to run but do not receive updates. To resolve the issue, try debugging your Satellite location or cluster.
registered Satellite Config components are installed in the cluster, but no resources are currently watched. To set up Watch-keeper, see Reviewing resources that are managed by Satellite Config.

Setting up Monitoring for Satellite location platform metrics

Forward and view additional metrics for Satellite in an IBM Cloud Monitoring instance that is enabled for platform-level metrics.

Metrics are available for the Satellite Link component of your location to help you monitor the performance of specific Link endpoints or of all Link endpoints for the location. For example, you can monitor the latency or throughput of a specific Link endpoint that you created.

  1. Create or choose an existing Monitoring instance.

    • If you already have a Monitoring instance in the same IBM Cloud region that your Satellite location is managed from, and the Monitoring instance is configured to collect platform metrics, the metrics that are generated for your Satellite location are automatically forwarded to this Monitoring instance.
    • Otherwise, to set up Monitoring for your Satellite location:
      1. Provision an IBM Cloud Monitoring instance in the same IBM Cloud region that your Satellite location is managed from.
      2. Enable the instance for platform-level metrics collection. Note that within one region, only one Monitoring instance can be enabled for platform metrics collection.

  2. In the Monitoring dashboard, click Open Dashboard for your Monitoring instance.

  3. In the Monitoring dashboard, click Dashboards > IBM > Satellite Link - Overview. The pre-defined dashboard for Satellite Link metrics opens. Note that if you just created this Monitoring instance, it might take up to two hours for the **IBM ** dashboards to become available.

    You can create a copy of this dashboard to customize the metrics that are shown. To add metrics that are enabled for IBM Cloud Satellite, search for the ibm_satellite_link prefix.

  4. Review the available metrics and attributes for segmentation.

  5. Review more ways that you can work with platform metrics.

After you set up Monitoring with the pre-defined dashboard for Satellite Link metrics, you can quickly access this dashboard from the Link endpoints tab of your Satellite location console by clicking Launch monitoring.

Available metrics

The following metrics are available for your Satellite location control plane.

Attributes for segmentation

Review the following global and additional attributes that are available for segmentation of Satellite metrics.

Global attributes

The following global attributes are available for segmenting all the available metrics.

Global attributes for metric segmentation
Attribute Attribute Name Attribute Description
Cloud Type ibm_ctype The cloud type, which can be public, dedicated, or local
Location ibm_location The location of the monitored resource, which can be a region, data center, or global
Resource ibm_resource The resource that is measured by the service, usually reported as an identifying name or GUID
Resource Type ibm_resource_type The type of the resource that is measured by the service
Resource group ibm_resource_group_name The resource group where the Satellite location was created
Scope ibm_scope The account GUID associated with this metric
Service name ibm_service_name The name of the service that generates this metric

Additional attributes

The following additional attributes that are specific to Satellite are available for segmenting one or more of the available metrics. See the Segment By field for each metric to determine its available segmentation attributes.

Additional attributes for metric segmentation
Attribute Attribute Name Attribute Description
Endpoint ID ibm_satellite_link_endpoint_id The identifier of the endpoint
Endpoint Name ibm_satellite_link_endpoint_name The name of the endpoint
Location ID ibm_satellite_link_location_id The identifier of the location
Service instance ibm_service_instance The service instance segment identifies the instance the metric is associated with
Service instance name ibm_service_instance_name The user-provided name of the service instance, which might not be unique across regions in the account

Setting up monitoring for clusters

You cannot currently use the Red Hat OpenShift on IBM Cloud console or the observability plug-in CLI (ibmcloud ob) to enable monitoring for Satellite clusters. You must manually deploy monitoring agents to your cluster to forward metrics to Monitoring.

To set up monitoring for Red Hat OpenShift clusters that run in your Satellite location, see Deploying a monitoring agent in a Red Hat OpenShift cluster. When you specify address for the COLLECTOR_ENDPOINT, you can use the satellite-sysdig link endpoint address so that you don't need to open up new firewall rules.