Service Telemetry Framework 1.0

Red Hat OpenStack Platform 16.0

Installing and deploying Service Telemetry Framework 1.0

OpenStack Documentation Team


This guide contains information about installing the core components and deploying Service Telemetry Framework 1.0.

Chapter 1. Introduction to Service Telemetry Framework

Service Telemetry Framework (STF) provides automated collection of measurements and data from remote clients - Red Hat OpenStack Platform or third-party nodes - and transmission of that information to a centralized, receiving Red Hat OpenShift Container Platform (OCP) deployment for storage, retrieval, and monitoring. The data can be either of two types:

a numeric measurement of an application or system
irregular and discrete occurrences that happen in a system

The collection components that are required on the clients are lightweight. The multicast message bus that is shared by all clients and the deployment provides fast and reliable data transport. Other modular components for receiving and storing data are deployed in containers on OCP.

STF provides access to monitoring functions such as alert generation, visualization through dashboards, and single source of truth telemetry analysis to support orchestration.

1.1. Service Telemetry Framework architecture

Service Telemetry Framework (STF) uses the components described in Table 1.1, “STF components”:

Table 1.1. STF components

ClientComponentServer (OCP)


An AMQP 1.x compatible messaging bus to shuttle the metrics to STF for storage in Prometheus



Smart Gateway to pick metrics and events from the AMQP 1.x bus and to deliver events to ElasticSearch or to provide metrics to Prometheus



Prometheus as time-series data storage



ElasticSearch as events data storage



collectd to collect infrastructure metrics and events



Ceilometer to collect Red Hat OpenStack Platform metrics and events


Figure 1.1. Service Telemetry Framework architecture overview

Service Telemetry Framework architecture overview

The Service Telemetry Framework data collection components, collectd and Ceilometer, and the transport components, AMQ Interconnect and Smart Gateway, are fully supported. The data storage components, Prometheus and ElasticSearch, including the Operator artifacts, and visualization component Grafana are community-supported, and are not officially supported.

For metrics, on the client side, collectd collects high-resolution metrics. collectd delivers the data to Prometheus by using the AMQP1 plugin, which places the data onto the message bus. On the server side, a Golang application called the Smart Gateway takes the data stream from the bus and exposes it as a local scrape endpoint for Prometheus.

If you plan to collect and store events, collectd or Ceilometer delivers event data to the server side by using the AMQP1 plugin, which places the data onto the message bus. Another Smart Gateway writes the data to the ElasticSearch datastore.

Server-side STF monitoring infrastructure consists of the following layers:

  • Service Telemetry Framework 1.0 (STF)
  • Red Hat OpenShift Container Platform (OCP)
  • Infrastructure platform

Figure 1.2. Server-side STF monitoring infrastructure

Server-side STF monitoring infrastructure

For more information about how to deploy Red Hat OpenShift Container Platform, see the OCP product documentation. You can install OCP on cloud platforms or on bare metal. For more information about STF performance and scaling, see


Do not install OCP on the same infrastructure that you want to monitor.

1.2. Installation size

The size of your Red Hat OpenShift Container Platform installation depends on the following factors:

  • The number of nodes you want to monitor.
  • The number of metrics you want to collect.
  • The resolution of metrics.
  • The length of time that you want to store the data.

Installation of Service Telemetry Framework (STF) depends on the existing Red Hat OpenShift Container Platform environment. Ensure that you install monitoring for Red Hat OpenStack Platform on a platform separate from your Red Hat OpenStack Platform environment. You can install Red Hat OpenShift Container Platform (OCP) on baremetal or other supported cloud platforms. For more information about installing OCP, see OpenShift Container Platform 4.3 Documentation.

The size of your OCP environment depends on the infrastructure you select. For more information about minimum resources requirements when installing OCP on baremetal, see Minimum resource requirements in the Installing a cluster on bare metal guide. For installation requirements of the various public and private cloud platforms which you can install, see the corresponding installation documentation for your cloud platform of choice.

Chapter 2. Installing the core components of Service Telemetry Framework

Before you install Service Telemetry Framework (STF), ensure that Red Hat OpenShift Container Platform (OCP) version 4.x is running and that you understand the core components of the framework. As part of the OCP installation planning process, ensure that the administrator provides persistent storage and enough resources to run the STF components on top of the OCP environment.


Red Hat OpenShift Container Platform version 4.3 or later is currently required for a successful installation of STF.

2.1. The core components of STF

The following STF core components are managed by Operators:

  • Prometheus and AlertManager
  • ElasticSearch
  • Smart Gateway
  • AMQ Interconnect

Each component has a corresponding Operator that you can use to load the various application components and objects.

Additional resources

For more information about Operators, see the Understanding Operators guide.

2.2. Preparing your OCP environment for STF

As you prepare your OCP environment for STF, you must plan for persistent storage, adequate resources, and event storage:

2.2.1. Persistent volumes

STF uses persistent storage in OCP to instantiate the volumes dynamically so that Prometheus and ElasticSearch can store metrics and events.

Additional resources

For more information about configuring persistent storage for OCP, see Understanding persistent storage. Using ephemeral storage


You can use ephemeral storage with STF. However, if you use ephemeral storage, you might experience data loss if a pod is restarted, updated, or rescheduled onto another node. Use ephemeral storage only for development or testing, and not production environments.


  • To enable ephemeral storage for STF, set storageEphemeralEnabled: true in your ServiceTelemetry manifest.

Additional resources

For more information about enabling ephemeral storage for STF, see Section 4.6.1, “Configuring ephemeral storage”.

2.2.2. Resource allocation

To enable the scheduling of pods within the OCP infrastructure, you need resources for the components that are running. If you do not allocate enough resources, pods remain in a Pending state because they cannot be scheduled.

The amount of resources that you require to run STF depends on your environment and the number of nodes and clouds that you want to monitor.

Additional resources

For recommendations about sizing for metrics collection see

For information about sizing requirements for ElasticSearch, see

2.2.3. Node tuning operator

STF uses ElasticSearch to store events, which requires a larger than normal vm.max_map_count. The vm.max_map_count value is set by default in Red Hat OpenShift Container Platform.

If you want to edit the value of vm.max_map_count, you cannot apply node tuning manually using the sysctl command because Red Hat OpenShift Container Platform manages nodes directly. To configure values and apply them to the infrastructure, you must use the node tuning operator. For more information, see Using the Node Tuning Operator.

In an OCP deployment, the default node tuning operator specification provides the required profiles for ElasticSearch workloads or pods scheduled on nodes. To view the default cluster node tuning specification, run the following command:

oc get Tuned/default -o yaml -n openshift-cluster-node-tuning-operator

The output of the default specification is documented at Default profiles set on a cluster. The assignment of profiles is managed in the recommend section where profiles are applied to a node when certain conditions are met. When scheduling ElasticSearch to a node in STF, one of the following profiles is applied:

  • openshift-control-plane-es
  • openshift-node-es

When scheduling an ElasticSearch pod, there must be a label present that matches If the label is present, one of the two profiles is assigned to the pod. No action is required by the administrator if you use the recommended Operator for ElasticSearch. If you use a custom-deployed ElasticSearch with STF, ensure that you add the label to all scheduled pods.

Additional resources

For more information about virtual memory usage by ElasticSearch, see

For more information about how the profiles are applied to nodes, see Custom tuning specification.

2.3. Deploying STF to the OCP environment

You can deploy STF to the OCP environment in one of two ways:

2.3.1. Deploying STF to the OCP environment with ElasticSearch

Complete the following tasks:

2.3.2. Deploying STF to the OCP environment without ElasticSearch

Complete the following tasks:

2.3.3. Creating a namespace

Create a namespace to hold the STF components. The service-telemetry namespace is used throughout the documentation:


  • Enter the following command:

    oc new-project service-telemetry

2.3.4. Creating an OperatorGroup

Create an OperatorGroup in the namespace so that you can schedule the Operator pods.


  • Enter the following command:

    oc apply -f - <<EOF
    kind: OperatorGroup
      name: service-telemetry-operator-group
      namespace: service-telemetry
      - service-telemetry

Additional resources

For more information, see OperatorGroups.

2.3.5. Enabling the Community Catalog Source

Before you install ElasticSearch, you must have access to the resources on the Community Catalog Source:


  • Enter the following command:

    oc apply -f - <<EOF
    kind: CatalogSource
      name: operatorhubio-operators
      namespace: openshift-marketplace
      sourceType: grpc
      displayName: Operators

2.3.6. Enabling Red Hat STF Operator Source

Before you deploy STF on Red Hat OpenShift Container Platform, you must enable the operator source.


  1. Install an OperatorSource that contains the Service Telemetry Operator and the Smart Gateway Operator:

    oc apply -f - <<EOF
    kind: OperatorSource
        opsrc-provider: redhat-operators-stf
      name: redhat-operators-stf
      namespace: openshift-marketplace
      authorizationToken: {}
      displayName: Red Hat STF Operators
      publisher: Red Hat
      registryNamespace: redhat-operators-stf
      type: appregistry
  2. To validate the creation of your OperatorSource, use the oc get operatorsources command. A successful import results in the MESSAGE field returning a result of The object has been successfully reconciled.

    $ oc get -nopenshift-marketplace operatorsource redhat-operators-stf
    NAME                   TYPE          ENDPOINT              REGISTRY               DISPLAYNAME             PUBLISHER   STATUS      MESSAGE
    redhat-operators-stf   appregistry   redhat-operators-stf   Red Hat STF Operators   Red Hat     Succeeded   The object has been successfully reconciled
  3. To validate that the Operators are available from the catalog, use the oc get packagemanifest command:

    $ oc get packagemanifests | grep "Red Hat STF"
    smartgateway-operator                        Red Hat STF Operators      2m50s
    servicetelemetry-operator                    Red Hat STF Operators      2m50s

2.3.7. Subscribing to the AMQ Certificate Manager Operator

You must subscribe to the AMQ Certificate Manager Operator before you deploy the other STF components because the AMQ Certificate Manager Operator runs globally-scoped and is not compatible with the dependency management of Operator Lifecycle Manager when used with other namespace-scoped operators.


  1. Subscribe to the AMQ Certificate Manager Operator, create the subscription, and validate the AMQ7 Certificate Manager:


    The AMQ Certificate Manager is installed globally for all namespaces, so the namespace value provided is openshift-operators. You might not see your amq7-cert-manager.v1.0.0 ClusterServiceVersion in the service-telemetry namespace for a few minutes until the processing executes against the namespace.

    oc apply -f - <<EOF
    kind: Subscription
      name: amq7-cert-manager
      namespace: openshift-operators
      channel: alpha
      installPlanApproval: Automatic
      name: amq7-cert-manager
      source: redhat-operators
      sourceNamespace: openshift-marketplace
  2. To validate your ClusterServiceVersion, use the oc get csv command. Ensure that amq7-cert-manager.v1.0.0 has a phase Succeeded.

    $ oc get --namespace openshift-operators csv
    NAME                       DISPLAY                                         VERSION   REPLACES   PHASE
    amq7-cert-manager.v1.0.0   Red Hat Integration - AMQ Certificate Manager   1.0.0                Succeeded

2.3.8. Subscribing to the Elastic Cloud on Kubernetes Operator

Before you install the Service Telemetry Operator and if you plan to store events in ElasticSearch, you must enable the Elastic Cloud Kubernetes Operator.


  1. Apply the following manifest to your OCP environment to enable the Elastic Cloud on Kubernetes Operator:

    oc apply -f - <<EOF
    kind: Subscription
      name: elastic-cloud-eck
      namespace: service-telemetry
      channel: stable
      installPlanApproval: Automatic
      name: elastic-cloud-eck
      source: operatorhubio-operators
      sourceNamespace: openshift-marketplace
  2. To verify that the ClusterServiceVersion for ElasticSearch Cloud on Kubernetes succeeded, enter the oc get csv command:

    $ oc get csv
    NAME                       DISPLAY                                         VERSION   REPLACES                   PHASE
    elastic-cloud-eck.v1.1.0   Elastic Cloud on Kubernetes                     1.1.0     elastic-cloud-eck.v1.0.1   Succeeded

2.3.9. Subscribing to the Service Telemetry Operator

To instantiate an STF instance, create the ServiceTelemetry object to allow the Service Telemetry Operator to create the environment.


  1. To create the Service Telemetry Operator subscription, enter the oc apply -f command:

    oc apply -f - <<EOF
    kind: Subscription
      name: servicetelemetry-operator
      namespace: service-telemetry
      channel: stable
      installPlanApproval: Automatic
      name: servicetelemetry-operator
      source: redhat-operators-stf
      sourceNamespace: openshift-marketplace
  2. To validate the Service Telemetry Operator and the dependent operators, enter the following command:

    $ oc get csv --namespace service-telemetry
    NAME                                DISPLAY                                         VERSION   REPLACES                            PHASE
    amq7-cert-manager.v1.0.0            Red Hat Integration - AMQ Certificate Manager   1.0.0                                         Succeeded
    amq7-interconnect-operator.v1.2.0   Red Hat Integration - AMQ Interconnect          1.2.0                                         Succeeded
    elastic-cloud-eck.v1.1.0            Elastic Cloud on Kubernetes                     1.1.0     elastic-cloud-eck.v1.0.1            Succeeded
    prometheusoperator.0.37.0           Prometheus Operator                             0.37.0    prometheusoperator.0.32.0           Succeeded
    service-telemetry-operator.v1.0.2   Service Telemetry Operator                      1.0.2     service-telemetry-operator.v1.0.1   Succeeded
    smart-gateway-operator.v1.0.1       Smart Gateway Operator                          1.0.1     smart-gateway-operator.v1.0.0       Succeeded

2.3.10. Creating a ServiceTelemetry object in OCP

To deploy the Service Telemetry Framework, you must create an instance of ServiceTelemetry in OCP. By default, eventsEnabled is set to false. If you do not want to store events in ElasticSearch, ensure that eventsEnabled is set to false. For more information, see Section 2.3.2, “Deploying STF to the OCP environment without ElasticSearch”.

The following core parameters are available for a ServiceTelemetry manifest:

Table 2.1. Core parameters for a ServiceTelemetry manifest

ParameterDescriptionDefault Value


Enable events support in STF. Requires prerequisite steps to ensure ElasticSearch can be started. For more information, see Section 2.3.8, “Subscribing to the Elastic Cloud on Kubernetes Operator”.



Enable metrics support in STF.



Enable high availability in STF. For more information, see Section 4.3, “High availability”.



Enable ephemeral storage support in STF. For more information, see Section 4.6, “Ephemeral storage”.



  1. To store events in ElasticSearch, set eventsEnabled to true during deployment:

    oc apply -f - <<EOF
    kind: ServiceTelemetry
      name: stf-default
      namespace: service-telemetry
      eventsEnabled: true
      metricsEnabled: true
  2. To view the STF deployment logs in the Service Telemetry Operator, use the oc logs command:

    oc logs $(oc get pod --selector='name=service-telemetry-operator' -oname) -c ansible
    PLAY RECAP ***
    localhost                  : ok=37   changed=0    unreachable=0    failed=0    skipped=1    rescued=0    ignored=0
  3. View the pods and the status of each pod to determine that all workloads are operating nominally:


    If you set eventsEnabled: true, the notification Smart Gateways will Error and CrashLoopBackOff for a period of time before ElasticSearch starts.

    $ oc get pods
    NAME                                                              READY   STATUS             RESTARTS   AGE
    alertmanager-stf-default-0                                        2/2     Running            0          26m
    elastic-operator-645dc8b8ff-jwnzt                                 1/1     Running            0          88m
    elasticsearch-es-default-0                                        1/1     Running            0          26m
    interconnect-operator-6fd49d9fb9-4bl92                            1/1     Running            0          46m
    prometheus-operator-bf7d97fb9-kwnlx                               1/1     Running            0          46m
    prometheus-stf-default-0                                          3/3     Running            0          26m
    service-telemetry-operator-54f4c99d9b-k7ll6                       2/2     Running            0          46m
    smart-gateway-operator-7ff58bcf94-66rvx                           2/2     Running            0          46m
    stf-default-ceilometer-notification-smartgateway-6675df547q4lbj   1/1     Running            0          26m
    stf-default-collectd-notification-smartgateway-698c87fbb7-xj528   1/1     Running            0          26m
    stf-default-collectd-telemetry-smartgateway-79c967c8f7-9hsqn      1/1     Running            0          26m
    stf-default-interconnect-7458fd4d69-nqbfs                         1/1     Running            0          26m

2.4. Removing STF from the OCP environment

Remove STF from an OCP environment if you no longer require the STF functionality.

Complete the following tasks:

2.4.1. Deleting the namespace

To remove the operational resources for STF from OCP, delete the namespace.


  1. Run the oc delete command:

    oc delete project service-telemetry
  2. Verify that the resources have been deleted from the namespace:

    $ oc get all
    No resources found.

2.4.2. Removing the OperatorSource

If you do not expect to install Service Telemetry Framework again, delete the OperatorSource. When you remove the OperatorSource, PackageManifests related to STF are removed from the Operator Lifecycle Manager catalog.


  1. Delete the OperatorSource:

    $ oc delete --namespace=openshift-marketplace operatorsource redhat-operators-stf "redhat-operators-stf" deleted
  2. Verify that the STF PackageManifests are removed from the platform. If successful, the following command returns no result:

    $ oc get packagemanifests | grep "Red Hat STF"
  3. If you enabled the Community Catalog Source during the installation process and you no longer need this catalog source, delete it:

    $ oc delete --namespace=openshift-marketplace catalogsource operatorhubio-operators "operatorhubio-operators" deleted

Additional resources

For more information about the Community Catalog Source, see Section 2.3, “Deploying STF to the OCP environment”.

Chapter 3. Completing the Service Telemetry Framework configuration

3.1. Connecting Red Hat OpenStack Platform to Service Telemetry Framework

To collect metrics, events, or both, and to send them to the Service Telemetry Framework (STF) storage domain, you must configure the Red Hat OpenStack Platform overcloud to enable data collection and transport.

To deploy data collection and transport to STF on Red Hat OpenStack Platform cloud nodes that employ routed L3 domains, such as distributed compute node (DCN) or spine-leaf, see Section 3.2, “Deploying to non-standard network topologies”.

3.2. Deploying to non-standard network topologies

If your nodes are on a separate network from the default InternalApi network, you must make configuration adjustments so that AMQ Interconnect can transport data to the Service Telemetry Framework (STF) server instance. This scenario is typical in a spine-leaf or a DCN topology. For more information about DCN configuration, see the Spine Leaf Networking guide.

If you use STF with Red Hat OpenStack Platform 16.0 and plan to monitor your Ceph, Block, or Object storage nodes, you must make configuration changes that are similar to the configuration changes that you make to the spine-leaf and DCN network configuration. To monitor Ceph nodes, use the CephStorageExtraConfig parameter to define which network interface to load into the AMQ Interconnect and collectd configuration files.

      tripleo::profile::base::metrics::collectd::amqp_host: "%{hiera('storage')}"
      tripleo::profile::base::metrics::qdr::listener_addr: "%{hiera('storage')}"
      tripleo::profile::base::ceilometer::agent::notification::notifier_host_addr: "%{hiera('storage')}"

Similarly, you must specify BlockStorageExtraConfig and ObjectStorageExtraConfig parameters if your environment uses Block and Object storage roles.

The deployment of a spine-leaf topology involves creating roles and networks, then assigning those networks to the available roles. When you configure data collection and transport for STF for an Red Hat OpenStack Platform deployment, the default network for roles is InternalApi. For Ceph, Block and Object storage roles, the default network is Storage. Because a spine-leaf configuration can result in different networks being assigned to different Leaf groupings and those names are typically unique, additional configuration is required in the parameter_defaults section of the Red Hat OpenStack Platform environment files.


  1. Document which networks are available for each of the Leaf roles. For examples of network name definitions, see Creating a network data file in the Spine Leaf Networking guide. For more information about the creation of the Leaf groupings (roles) and assignment of the networks to those groupings, see Creating a roles data file in the Spine Leaf Networking guide.
  2. Add the following configuration example to the ExtraConfig section for each of the leaf roles. In this example, internal_api_subnet is the value defined in the name_lower parameter of your network definition (with _subnet appended to the name for Leaf 0) , and is the network to which the ComputeLeaf0 leaf role is connected. In this case, the network identification of 0 corresponds to the Compute role for leaf 0, and represents a value that is different from the default internal API network name.

    For the ComputeLeaf0 leaf role, specify extra configuration to perform a hiera lookup to determine which network interface for a particular network to assign to the collectd AMQP host parameter. Perform the same configuration for the AMQ Interconnect listener address parameter.

    ›   tripleo::profile::base::metrics::collectd::amqp_host: "%{hiera('internal_api_subnet')}"
    ›   tripleo::profile::base::metrics::qdr::listener_addr: "%{hiera('internal_api_subnet')}"

    Additional leaf roles typically replace _subnet with _leafN where N represents a unique indentifier for the leaf.

    ›   tripleo::profile::base::metrics::collectd::amqp_host: "%{hiera('internal_api_leaf1')}"
    ›   tripleo::profile::base::metrics::qdr::listener_addr: "%{hiera('internal_api_leaf1')}"

    This example configuration is on a CephStorage leaf role:

    ›   tripleo::profile::base::metrics::collectd::amqp_host: "%{hiera('storage_subnet')}"
    ›   tripleo::profile::base::metrics::qdr::listener_addr: "%{hiera('storage_subnet')}"

3.3. Configuring Red Hat OpenStack Platform overcloud for Service Telemetry Framework

To configure the Red Hat OpenStack Platform overcloud, you must configure the data collection applications and the data transport to STF, and deploy the overcloud.

To configure the Red Hat OpenStack Platform overcloud, complete the following tasks:

3.3.1. Retrieving the AMQ Interconnect route address

When you configure the Red Hat OpenStack Platform overcloud for STF, you must provide the AMQ Interconnect route address in the STF connection file.


  1. Log in to your Red Hat OpenShift Container Platform (OCP) environment.
  2. In the service-telemetry project, retrieve the AMQ Interconnect route address:

    $ oc get routes -ogo-template='{{ range .items }}{{printf "%s\n" }}{{ end }}' | grep "\-5671"

    If your STF installation differs from the documentation, ensure that you retrieve the correct AMQ Interconnect route address.

3.3.2. Configuring the STF connection for the overcloud

To configure the STF connection, you must create a file that contains the connection configuration of the AMQ Interconnect for the overcloud to the STF deployment. Enable the collection of events and storage of the events in STF and deploy the overcloud.


  1. Log in to the Red Hat OpenStack Platform undercloud as the stack user.
  2. Create a configuration file called stf-connectors.yaml in the /home/stack directory.


    The Service Telemetry Operator simplifies the deployment of all data ingestion and data storage components for single cloud deployments. To share the data storage domain with multiple clouds, see Section 4.5, “Configuring multiple clouds”.

  3. In the stf-connectors.yaml file, configure the MetricsQdrConnectors address to connect the AMQ Interconnect on the overcloud to the STF deployment.

    • Add the CeilometerQdrPublishEvents: true parameter to enable collection and transport of Ceilometer events to STF.
    • Replace the host parameter with the value of HOST/PORT that you retrieved in Section 3.3.1, “Retrieving the AMQ Interconnect route address”:

          EventPipelinePublishers: []
          CeilometerQdrPublishEvents: true
          - host:
            port: 443
            role: edge
            sslProfile: sslProfile
            verifyHostname: false
  4. Add the following files to your Red Hat OpenStack Platform director deployment to setup collectd and AMQ Interconnect:

    • the stf-connectors.yaml environment file
    • the enable-stf.yaml file that ensures that the environment is being used during the overcloud deployment
    • the ceilometer-write-qdr.yaml file that ensures that Ceilometer telemetry is sent to STF

      openstack overcloud deploy <other arguments>
        --templates /usr/share/openstack-tripleo-heat-templates \
        --environment-file <...other-environment-files...> \
        --environment-file /usr/share/openstack-tripleo-heat-templates/environments/metrics/ceilometer-write-qdr.yaml \
        --environment-file /usr/share/openstack-tripleo-heat-templates/environments/enable-stf.yaml \
        --environment-file /home/stack/stf-connectors.yaml
  5. Deploy the Red Hat OpenStack Platform overcloud.

3.3.3. Validating client-side installation

To validate data collection from the STF storage domain, query the data sources for delivered data. To validate individual nodes in the Red Hat OpenStack Platform deployment, connect to the console using SSH.


  1. Log in to an overcloud node, for example, controller-0.
  2. Ensure that metrics_qdr container is running on the node:

    $ sudo podman container inspect --format '{{.State.Status}}' metrics_qdr
  3. Return the internal network address on which AMQ Interconnect is running, for example, listening on port 5666:

    $ sudo podman exec -it metrics_qdr cat /etc/qpid-dispatch/qdrouterd.conf
    listener {
        port: 5666
        authenticatePeer: no
        saslMechanisms: ANONYMOUS
  4. Return a list of connections to the local AMQ Interconnect:

    $ sudo podman exec -it metrics_qdr qdstat --bus= --connections
      id   host                                                                  container                                                                                                  role    dir  security                            authentication  tenant
      1  stf-default-interconnect-7458fd4d69-bgzfb                                                                  edge    out  TLSv1.2(DHE-RSA-AES256-GCM-SHA384)  anonymous-user
      12                                             normal  in   no-security                         no-auth
      16                                                     metrics                                                                                                    normal  in   no-security                         anonymous-user
      899                                                     10a2e99d-1b8a-4329-b48c-4335e5f75c84                                                                       normal  in   no-security                         no-auth

    There are four connections:

    • Outbound connection to STF
    • Inbound connection from collectd
    • Inbound connection from ceilometer
    • Inbound connection from our qdstat client

      The outbound STF connection is provided to the MetricsQdrConnectors host parameter and is the route for the STF storage domain. The other hosts are internal network addresses of the client connections to this AMQ Interconnect.

  5. To ensure that messages are being delivered, list the links, and view the _edge address in the deliv column for delivery of messages:

    $ sudo podman exec -it metrics_qdr qdstat --bus= --links
    Router Links
      type      dir  conn id  id    peer  class   addr                  phs  cap  pri  undel  unsett  deliv    presett  psdrop  acc  rej  rel  mod  delay  rate
      endpoint  out  1        5           local   _edge                      250  0    0      0       2979926  2979924  0       0    0    2    0    0      0
      endpoint  in   1        6                                              250  0    0      0       0        0        0       0    0    0    0    0      0
      endpoint  in   1        7                                              250  0    0      0       0        0        0       0    0    0    0    0      0
      endpoint  out  1        8                                              250  0    0      0       0        0        0       0    0    0    0    0      0
      endpoint  in   1        9                                              250  0    0      0       0        0        0       0    0    0    0    0      0
      endpoint  out  1        10                                             250  0    0      0       911      911      0       0    0    0    0    911    0
      endpoint  in   1        11                                             250  0    0      0       0        911      0       0    0    0    0    0      0
      endpoint  out  12       32          local   temp.lSY6Mcicol4J2Kp       250  0    0      0       0        0        0       0    0    0    0    0      0
      endpoint  in   16       41                                             250  0    0      0       2979924  2979924  0       0    0    0    0    0      0
      endpoint  in   912      1834        mobile  $management           0    250  0    0      0       1        0        0       1    0    0    0    0      0
      endpoint  out  912      1835        local   temp.9Ok2resI9tmt+CT       250  0    0      0       0        0        0       0    0    0    0    0      0
  6. To list the addresses from Red Hat OpenStack Platform nodes to STF, connect to OCP to get the AMQ Interconnect pod name and list the connections. List the available AMQ Interconnect pods:

    $ oc get pods -l application=stf-default-interconnect
    NAME                                        READY   STATUS    RESTARTS   AGE
    stf-default-interconnect-7458fd4d69-bgzfb   1/1     Running   0          6d21h
  7. Connect to the pod and run the qdstat --connections command to list the known connections:

    $ oc exec -it stf-default-interconnect-7458fd4d69-bgzfb -- qdstat --connections
    2020-04-21 18:25:47.243852 UTC
      id  host               container                                                             role    dir  security                                authentication  project  last dlv      uptime
      1  rcv[stf-default-collectd-telemetry-smartgateway-79c967c8f7-kq4qv]     normal  in   no-security                             anonymous-user          000:00:00:00  006:21:50:25
      2  rcv[stf-default-ceilometer-notification-smartgateway-6675df547mbjk5]  normal  in   no-security                             anonymous-user          000:21:25:57  006:21:49:36
      3  rcv[stf-default-collectd-notification-smartgateway-698c87fbb7-f28v6]  normal  in   no-security                             anonymous-user          000:21:36:53  006:21:49:09
      22   Router.ceph-0.redhat.local                                            edge    in   TLSv1/SSLv3(DHE-RSA-AES256-GCM-SHA384)  anonymous-user          000:00:00:03  000:22:08:43
      23   Router.compute-0.redhat.local                                         edge    in   TLSv1/SSLv3(DHE-RSA-AES256-GCM-SHA384)  anonymous-user          000:00:00:03  000:22:08:43
      24   Router.controller-0.redhat.local                                      edge    in   TLSv1/SSLv3(DHE-RSA-AES256-GCM-SHA384)  anonymous-user          000:00:00:00  000:22:08:34
      27    c2f541c1-4c97-4b37-a189-a396c08fb079                                  normal  in   no-security                             no-auth                 000:00:00:00  000:00:00:00

    In this example, there are three edge connections from the Red Hat OpenStack Platform nodes with connection id 22, 23, and 24.

  8. To view the number of messages delivered by the network, use each address with the oc exec command:

    $ oc exec -it stf-default-interconnect-7458fd4d69-bgzfb -- qdstat --address
    2020-04-21 18:20:10.293258 UTC
    Router Addresses
      class   addr                              phs  distrib    pri  local  remote  in         out        thru  fallback
      mobile  anycast/ceilometer/event.sample   0    balanced   -    1      0       1,553      1,553      0     0
      mobile  collectd/notify                   0    multicast  -    1      0       10         10         0     0
      mobile  collectd/telemetry                0    multicast  -    1      0       7,798,049  7,798,049  0     0

Chapter 4. Advanced features

The following optional features can provide additional functionality to the Service Telemetry Framework (STF):

4.1. Customizing the deployment

The Service Telemetry Operator watches for a ServiceTelemetry manifest to load into Red Hat OpenShift Container Platform (OCP). The Operator then creates other objects in memory, which results in the dependent Operators creating the workloads they are responsible for managing.


When you override the manifest, you must provide the entire manifest contents, including object names or namespaces. There is no dynamic parameter substitution when you override a manifest.

To override a manifest successfully with Service Telemetry Framework (STF), deploy a default environment using the core options only. For more information about the core options, see Section 2.3.10, “Creating a ServiceTelemetry object in OCP”. When you deploy STF, use the oc get command to retrieve the default deployed manifest. When you use a manifest that was originally generated by Service Telemetry Operator, the manifest is compatible with the other objects that are managed by the Operators.

For example, when the metricsEnabled: true parameter is configured in the ServiceTelemetry manifest, the Service Telemetry Operator requests components for metrics retrieval and storage using the default manifests. In some cases, you might want to override the default manifest. For more information, see Section 4.1.1, “Manifest override parameters”.

4.1.1. Manifest override parameters

This table describes the available parameters that you can use to override a manifest, along with the corresponding retrieval commands.

Table 4.1. Manifest override parameters

Override parameterDescriptionRetrieval command


Override the Alertmanager object creation. The Prometheus Operator watches for Alertmanager objects.

oc get alertmanager stf-default -oyaml


Override the Secret that contains the Alertmanager configuration. The Prometheus Operator uses a secret named alertmanager-{{ alertmanager-name }}, for example, stf-default, to provide the alertmanager.yaml configuration to Alertmanager.

oc get secret alertmanager-stf-default -oyaml


Override the ElasticSearch object creation. The Elastic Cloud on Kuberneters Operator watches for ElasticSearch objects.

oc get elasticsearch elasticsearch -oyaml


Override the Interconnect object creation. The AMQ Interconnect Operator watches for Interconnect objects.

oc get interconnect stf-default-interconnect -oyaml


Override the Prometheus object creation. The Prometheus Operator watches for Prometheus objects.

oc get prometheus stf-default -oyaml


Override the ServiceMonitor object creation. The Prometheus Operator watches for ServiceMonitor objects.

oc get servicemonitor stf-default -oyaml


Override the SmartGateway object creation for collectd metrics. The Smart Gateway Operator watches for SmartGateway objects.

oc get smartgateway stf-default-collectd-telemetry -oyaml


Override the SmartGateway object creation for collectd events. The Smart Gateway Operator watches for SmartGateway objects.

oc get smartgateway stf-default-collectd-notification -oyaml


Override the SmartGateway object creation for Ceilometer events. The Smart Gateway Operator watches for SmartGateway objects.

oc get smartgateway stf-default-ceilometer-notification -oyaml

4.1.2. Overriding a managed manifest

Edit the ServiceTelemetry object and provide a parameter and manifest. For a list of available manifest override parameters, see Section 4.1, “Customizing the deployment”. The default ServiceTelemetry object is stf-default. Use oc get servicetelemetry to list the available STF deployments.


The oc edit command loads the default system editor. To override the default editor, pass or set the environment variable EDITOR to the preferred editor. For example, EDITOR=nano oc edit servicetelemetry stf-default.

  1. Log in to Red Hat OpenShift Container Platform.
  2. Change to the service-telemetry namespace:

    oc project service-telemetry
  3. Load the ServiceTelemetry object into an editor:

    oc edit servicetelemetry stf-default
  4. To modify the ServiceTelemetry object, provide a manifest override parameter and the contents of the manifest to write to OCP instead of the defaults provided by STF.


    The trailing pipe (|) after entering the manifest override parameter indicates that the value provided is multi-line.

    $ oc edit servicetelemetry stf-default
    kind: ServiceTelemetry
      annotations: |
      creationTimestamp: "2020-04-14T20:29:42Z"
      generation: 1
      name: stf-default
      namespace: service-telemetry
      resourceVersion: "1949423"
      selfLink: /apis/
      uid: d058bc41-1bb0-49f5-9a8b-642f4b8adb95
      metricsEnabled: true
      smartgatewayCollectdMetricsManifest: | 1
        kind: SmartGateway
          name: stf-default-collectd-telemetry
          namespace: service-telemetry
          amqpUrl: stf-default-interconnect.service-telemetry.svc.cluster.local:5672/collectd/telemetry
          debug: true
          prefetch: 15000
          serviceType: metrics
          size: 1
          useTimestamp: true 2
      - ansibleResult:
          changed: 0
          completion: 2020-04-14T20:32:19.079508
          failures: 0
          ok: 52
          skipped: 1
        lastTransitionTime: "2020-04-14T20:29:59Z"
        message: Awaiting next reconciliation
        reason: Successful
        status: "True"
        type: Running
    Manifest override parameter is defined in the spec of the ServiceTelemetry object.
    End of the manifest override content.
  5. Save and close.

4.2. Alerts

You create alert rules in Prometheus and alert routes in Alertmanager. Alert rules in Prometheus servers send alerts to an Alertmanager, which manages the alerts. Alertmanager can silence, inhibit, or aggregate alerts, and send notifications using email, on-call notification systems, or chat platforms.

To create an alert, complete the following tasks:

  1. Create an alert rule in Prometheus. For more information, see Section 4.2.1, “Creating an alert rule in Prometheus”.
  2. Create an alert route in Alertmanager. For more information, see Section 4.2.3, “Creating an alert route in Alertmanager”.

Additional resources

For more information about alerts or notifications with Prometheus and Alertmanager, see

To view an example set of alerts that you can use with Service Telemetry Framework (STF), see

4.2.1. Creating an alert rule in Prometheus

Prometheus evaluates alert rules to trigger notifications. If the rule condition returns an empty result set, the condition is false. Otherwise, the rule is true and it triggers an alert.

  1. Log in to Red Hat OpenShift Container Platform.
  2. Change to the service-telemetry namespace:

    oc project service-telemetry
  3. Create a PrometheusRule object that contains the alert rule. The Prometheus Operator loads the rule into Prometheus:

    oc apply -f - <<EOF
    kind: PrometheusRule
      creationTimestamp: null
        prometheus: stf-default
        role: alert-rules
      name: prometheus-alarm-rules
      namespace: service-telemetry
        - name: ./openstack.rules
            - alert: Metric Listener down
              expr: collectd_qpid_router_status < 1 # To change the rule, edit the value of the expr parameter.
  4. To verify that the rules have been loaded into Prometheus by the Operator, create a pod with access to curl:

    oc run curl --generator=run-pod/v1 --image=radial/busyboxplus:curl -i --tty
  5. Run curl to access the prometheus-operated service to return the rules loaded into memory:

    [ root@curl:/ ]$ curl prometheus-operated:9090/api/v1/rules
    {"status":"success","data":{"groups":[{"name":"./openstack.rules","file":"/etc/prometheus/rules/prometheus-stf-default-rulefiles-0/service-telemetry-prometheus-alarm-rules.yaml","rules":[{"name":"Metric Listener down","query":"collectd_qpid_router_status \u003c 1","duration":0,"labels":{},"annotations":{},"alerts":[],"health":"ok","type":"alerting"}],"interval":30}]}}
  6. To verify that the output shows the rules loaded into the PrometheusRule object, for example the output contains the defined ./openstack.rules, exit from the pod:

    [ root@curl:/ ]$ exit
  7. Clean up the environment by deleting the curl pod:

    $ oc delete pod curl
    pod "curl" deleted
Additional resources

For more information on alerting, see

4.2.2. Configuring custom alerts

You can add custom alerts to the PrometheusRule object that you created in Section 4.2.1, “Creating an alert rule in Prometheus”.

  1. Use the oc edit command:

    oc edit prometheusrules prometheus-alarm-rules
  2. Edit the PrometheusRules manifest.
  3. Save and close.
Additional resources

For more information about configuring alerting rules, see

For more information about PrometheusRules objects, see

4.2.3. Creating an alert route in Alertmanager

Use Alertmanager to deliver alerts to an external system, such as email, IRC, or other notification channel. The Prometheus Operator manages the Alertmanager configuration as an Red Hat OpenShift Container Platform (OCP) secret. STF by default deploys a basic configuration that results in no receivers:

alertmanager.yaml: |-
    resolve_timeout: 5m
    group_by: ['job']
    group_wait: 30s
    group_interval: 5m
    repeat_interval: 12h
    receiver: 'null'
  - name: 'null'

To deploy a custom Alertmanager route with STF, an alertmanagerConfigManifest parameter must be passed to the Service Telemetry Operator that results in an updated secret, managed by the Prometheus Operator. For more information, see Section 4.1.2, “Overriding a managed manifest”.

  1. Log in to Red Hat OpenShift Container Platform.
  2. Change to the service-telemetry namespace:

    oc project service-telemetry
  3. Edit the ServiceTelemetry object for your STF deployment

    oc edit servicetelemetry stf-default
  4. Add a new parameter, alertmanagerConfigManifest and the Secret object contents to define the alertmanager.yaml configuration for Alertmanager:


    This loads the default template that is already managed by Service Telemetry Operator. To validate the changes are populating correctly, change a value, return the alertmanager-stf-default secret, and verify that the new value is loaded into memory, for example, changing the value global.resolve_timeout from 5m to 10m.

    kind: ServiceTelemetry
      name: stf-default
      namespace: service-telemetry
      metricsEnabled: true
      alertmanagerConfigManifest: |
        apiVersion: v1
        kind: Secret
          name: 'alertmanager-stf-default'
          namespace: 'service-telemetry'
        type: Opaque
          alertmanager.yaml: |-
              resolve_timeout: 10m
              group_by: ['job']
              group_wait: 30s
              group_interval: 5m
              repeat_interval: 12h
              receiver: 'null'
            - name: 'null'
  5. Verify that the configuration was applied to the secret:

    $ oc get secret alertmanager-stf-default -o go-template='{{index .data "alertmanager.yaml" | base64decode }}'
      resolve_timeout: 10m
      group_by: ['job']
      group_wait: 30s
      group_interval: 5m
      repeat_interval: 12h
      receiver: 'null'
    - name: 'null'
  6. To verify the configuration has been loaded into Alertmanager, create a pod with access to curl:

    oc run curl --generator=run-pod/v1 --image=radial/busyboxplus:curl -i --tty
  7. Run curl against the alertmanager-operated service to retrieve the status and configYAML contents and review the supplied configuration matches the configuration loaded into Alertmanager:

    [ root@curl:/ ]$ curl alertmanager-operated:9093/api/v1/status
    {"status":"success","data":{"configYAML":"global:\n  resolve_timeout: 10m\n  http_config: {}\n  smtp_hello: localhost\n  smtp_require_tls: true\n  pagerduty_url:\n  hipchat_api_url:\n  opsgenie_api_url:\n  wechat_api_url:\n  victorops_api_url:\nroute:\n  receiver: \"null\"\n  group_by:\n  - job\n  group_wait: 30s\n  group_interval: 5m\n  repeat_interval: 12h\nreceivers:\n- name: \"null\"\ntemplates: []\n",...}}
  8. Verify that the configYAML field contains the expected changes. Exit from the pod:

    [ root@curl:/ ]$ exit
  9. To clean up the environment, delete the curl pod:

    $ oc delete pod curl
    pod "curl" deleted
Additional resources

For more information about the Red Hat OpenShift Container Platform secret and the Prometheus operator, see

4.3. High availability

High availability is the ability of Service Telemetry Framework (STF) to rapidly recover from failures in its component services. Although Red Hat OpenShift Container Platform (OCP) restarts a failed pod if nodes are available to schedule the workload, this recovery process might take more than one minute, during which time events and metrics are lost. A high availability configuration includes multiple copies of STF components, reducing recovery time to approximately 2 seconds. To protect against failure of an OCP node, deploy STF to an OCP cluster with three or more nodes.


STF is not yet a fully fault tolerant system. Delivery of metrics and events during the recovery period is not guaranteed.

Enabling high availability has the following effects:

  • Two AMQ Interconnect pods run instead of the default 1.
  • Three ElasticSearch pods run instead of the default 1.
  • Recovery time from a lost pod in either of these services reduces to approximately 2 seconds.

4.3.1. Configuring high availability

To configure STF for high availability, add highAvailabilityEnabled: true to the ServiceTelemetry object in OCP. You can this set this parameter at installation time or, if you already deployed STF, complete the following steps:

  1. Log in to Red Hat OpenShift Container Platform.
  2. Change to the service-telemetry namespace:

    oc project service-telemetry
  3. Use the oc command to edit the ServiceTelemetry object:

    $ oc edit ServiceTelemetry
  4. Add highAvailabilityEnabled: true to the spec section:

      eventsEnabled: true
      metricsEnabled: true
      highAvailabilityEnabled: true
  5. Save your changes and close the object.

4.4. Dashboards

Use third-party application Grafana to visualize system-level metrics gathered by collectd for each individual host node. For more information about configuring collectd, see Section 3.3, “Configuring Red Hat OpenStack Platform overcloud for Service Telemetry Framework”.

4.4.1. Setting up Grafana to host the dashboard

Grafana is not included in the default Service Telemetry Framework (STF) deployment so you must deploy the Grafana Operator from


  1. Log in to Red Hat OpenShift Container Platform.
  2. Change to the service-telemetry namespace:

    oc project service-telemetry
  3. Clone the dashboard repository.

    git clone
    cd dashboards
  4. Deploy the Grafana operator:

    oc create -f deploy/subscription.yaml
  5. To verify that the operator launched successfully, run the oc get csv command. If the value of the PHASE column is Succeeded, the operator launched successfully:

    $ oc get csv
    NAME                                DISPLAY                                         VERSION   REPLACES                            PHASE
    grafana-operator.v3.2.0             Grafana Operator                                3.2.0                                         Succeeded
  6. Launch a Grafana instance:

    $ oc create -f deploy/grafana.yaml
  7. Verify that the Grafana instance deployed:

    $ oc get pod -l app=grafana
    NAME                                  READY   STATUS    RESTARTS   AGE
    grafana-deployment-7fc7848b56-sbkhv   1/1     Running   0          1m
  8. Create the datasource and dashboard resources:

    oc create -f deploy/datasource.yaml \
        -f deploy/rhos-dashboard.yaml
  9. Verify that the resources installed correctly:

    $ oc get grafanadashboards
    NAME             AGE
    rhos-dashboard   7d21h
    $ oc get grafanadatasources
    NAME                                  AGE
    service-telemetry-grafanadatasource   1m
  10. Navigate to https://<grafana-route-address> in a web browser. Use the oc get routes command to retrieve the Grafana route address:

    oc get routes
  11. To view the dashboard, click Dashboards and Manage.

Additional resources Viewing and editing queries


  1. Log in to Red Hat OpenShift Container Platform. To view and edit queries, log in as the admin user.
  2. Change to the service-telemetry namespace:

    oc project service-telemetry
  3. To retrieve the default username and password, describe the Grafana object using oc describe:

    oc describe grafana service-telemetry-grafana

4.4.2. The Grafana infrastructure dashboard

The infrastructure dashboard shows metrics for a single node at a time. Select a node from the upper left corner of the dashboard. Top panels




Current Global Alerts


Current alerts fired by Prometheus

Recent Global Alerts


Recently fired alerts in 5m time steps

Status Panel


Node status: up, down, unavailable



Total operational time of node

CPU Cores


Total number of cores



Total memory

Disk Size


Total storage size



Total number of processes listed by type

Load Average


Load average represents the average number of running and uninterruptible processes residing in the kernel execution queue. Networking panels

Panels that display the network interfaces of the node.




Physical Interfaces Ingress Errors


Total errors with incoming data

Physical Interfaces Egress Errors


Total errors with outgoing data

Physical Interfaces Ingress Error Rates


Rate of incoming data errors

Physical Interfaces egress Error Rates


Rate of outgoing data errors

Physical Interfaces Packets Ingress pps Incoming packets per second

Physical Interfaces Packets Egress


Outgoing packets per second

Physical Interfaces Data Ingress


Incoming data rates

Physical Interfaces Data Egress


Outgoing data rates

Physical Interfaces Drop Rate Ingress


Incoming packets drop rate

Physical Interfaces Drop Rate Egress

pps CPU panels

Panels that display CPU usage of the node.


Current CPU Usage


Instantaneous usage at the time of the last query.

Aggregate CPU Usage


Average non-idle CPU activity of all cores on a node.

Aggr. CPU Usage by Type


Shows time spent for each type of thread averaged across all cores. Memory panels

Panels that display memory usage on the node.


Memory Used


Amount of memory being used at time of last query.

Huge Pages Used


Number of hugepages being used.

Memory Disk/file system

Panels that display space used on disk.


Disk Space Usage


Total disk use at time of last query.


Inode Usage


Total inode use at time of last query.


Aggregate Disk Space Usage


Total disk space used and reserved.

Because this query relies on the df plugin, temporary file systems that do not necessarily use disk space are included in the results. The query tries to filter out most of these, but it might not be exhaustive.

Disk Traffic


Shows rates for both reading and writing.


Disk Load


Approximate percentage of total disk bandwidth being used. The weighted I/O time series includes the backlog that might be accumulating. For more information, see the collectd disk plugin docs.




Operations done per second


Average I/O Operation Time


Average time each I/O operation took to complete. This average is not accurate, see the collectd disk plugin docs.


4.5. Configuring multiple clouds

You can configure multiple Red Hat OpenStack Platform clouds to target a single instance of Service Telemetry Framework (STF):

  1. Plan the AMQP address prefixes that you want to use for each cloud. For more information, see Section 4.5.1, “Planning AMQP address prefixes”.
  2. Deploy metrics and events consumer Smart Gateways for each cloud to listen on the corresponding address prefixes. For more information, see Section 4.5.2, “Deploying Smart Gateways”.
  3. Configure each cloud to send its metrics and events to STF on the correct address. For more information, see Section 4.5.3, “Creating the OpenStack environment file”.

Figure 4.1. Two Red Hat OpenStack Platform clouds connect to STF

OpenStack STF Overview 37 0919 topology

4.5.1. Planning AMQP address prefixes

By default, Red Hat OpenStack Platform nodes get data through two data collectors; collectd and Ceilometer. These components send telemetry data or notifications to the respective AMQP addresses, for example, collectd/telemetry, where STF Smart Gateways listen on those addresses for monitoring data.

To support multiple clouds and to identify which cloud generated the monitoring data, configure each cloud to send data to a unique address. Prefix a cloud identifier to the second part of the address. The following list shows some example addresses and identifiers:

  • collectd/cloud1-telemetry
  • collectd/cloud1-notify
  • anycast/ceilometer/cloud1-event.sample
  • collectd/cloud2-telemetry
  • collectd/cloud2-notify
  • anycast/ceilometer/cloud2-event.sample
  • collectd/us-east-1-telemetry
  • collectd/us-west-3-telemetry

4.5.2. Deploying Smart Gateways

You must deploy a Smart Gateway for each of the data collection types for each cloud; one for collectd metrics, one for collectd events, and one for Ceilometer events. Configure each of the Smart Gateways to listen on the AMQP address that you define for the corresponding cloud.

When you deploy STF for the first time, Smart Gateway manifests are created that define the initial Smart Gateways for a single cloud. When deploying Smart Gateways for multiple cloud support, you deploy multiple Smart Gateways for each of the data collection types that handle the metrics and the events data for each cloud. The initial Smart Gateways act as a template to create additional Smart Gateways, along with any authentication information required to connect to the data stores.

  1. Log in to Red Hat OpenShift Container Platform.
  2. Change to the service-telemetry namespace:

    oc project service-telemetry
  3. Use the initially deployed Smart Gateways as a template for additional Smart Gateways. List the currently deployed Smart Gateways with the oc get smartgateways command. For example, if you deployed STF with metricsEnabled: true and eventsEnabled: true, the following Smart Gateways are displayed in the output:

    $ oc get smartgateways
    NAME                                         AGE
    stf-default-ceilometer-notification          14d
    stf-default-collectd-notification            14d
    stf-default-collectd-telemetry               14d
  4. Retrieve the manifests for each Smart Gateway and store the contents in a temporary file, which you can modify later and use to create the new set of Smart Gateways:

    truncate --size 0 /tmp/cloud1-smartgateways.yaml && \
    for sg in $(oc get smartgateways -oname)
      echo "---" >> /tmp/cloud1-smartgateways.yaml
      oc get ${sg} -oyaml --export >> /tmp/cloud1-smartgateways.yaml
  5. Modify the Smart Gateway manifest in the /tmp/cloud1-smartgateways.yaml file. Adjust the and spec.amqpUrl fields to include the cloud identifier from your schema. For more information, see ]. To view example Smart Gateway manifests, see <<example-manifests_advanced-features[.
  6. Deploy your new Smart Gateways:

    oc apply -f /tmp/cloud1-smartgateways.yaml
  7. Verify that each Smart Gateway is running. This can take several minutes depending on the number of Smart Gateways:

    oc get po -l app=smart-gateway Example manifests


The content in the following examples might be different to the file content in your deployment. Copy the manifests in your deployment.

Ensure that the name and amqpUrl parameters of each Smart Gateway match the names that you want to use for your clouds. For more information, see Section 4.5.1, “Planning AMQP address prefixes”.


Your output may have some additional metadata parameters that you can remove from the manifests you that load into OCP.

kind: SmartGateway
  name: stf-default-ceilometer-notification-cloud1  1
  amqpDataSource: ceilometer
  amqpUrl: stf-default-interconnect.service-telemetry.svc.cluster.local:5672/anycast/ceilometer/cloud1-event.sample  2
  debug: false
  elasticPass: fkzfhghw......
  elasticUrl: https://elasticsearch-es-http.service-telemetry.svc.cluster.local:9200
  elasticUser: elastic
  resetIndex: false
  serviceType: events
  size: 1
  tlsCaCert: /config/certs/ca.crt
  tlsClientCert: /config/certs/tls.crt
  tlsClientKey: /config/certs/tls.key
  tlsServerName: elasticsearch-es-http.service-telemetry.svc.cluster.local
  useBasicAuth: true
  useTls: true
kind: SmartGateway
  name: stf-default-collectd-notification-cloud1  3
  amqpDataSource: collectd
  amqpUrl: stf-default-interconnect.service-telemetry.svc.cluster.local:5672/collectd/cloud1-notify  4
  debug: false
  elasticPass: fkzfhghw......
  elasticUrl: https://elasticsearch-es-http.service-telemetry.svc.cluster.local:9200
  elasticUser: elastic
  resetIndex: false
  serviceType: events
  size: 1
  tlsCaCert: /config/certs/ca.crt
  tlsClientCert: /config/certs/tls.crt
  tlsClientKey: /config/certs/tls.key
  tlsServerName: elasticsearch-es-http.service-telemetry.svc.cluster.local
  useBasicAuth: true
  useTls: true
kind: SmartGateway
  name: stf-default-collectd-telemetry-cloud1 5
  amqpUrl: stf-default-interconnect.service-telemetry.svc.cluster.local:5672/collectd/cloud1-telemetry  6
  debug: false
  prefetch: 15000
  serviceType: metrics
  size: 1
  useTimestamp: true
Name for Ceilometer notifications for cloud1
AMQP Address for Ceilometer notifications for cloud1
Name for collectd telemetry for cloud1
AMQP Address for collectd telemetry for cloud1
Name for collectd notifications for cloud1
AMQP Address for collectd notifications for cloud1

4.5.3. Creating the OpenStack environment file

To label traffic according to the cloud of origin, you must create a configuration with cloud-specific instance names. Create an stf-connectors.yaml file and adjust the values of CeilometerQdrEventsConfig and CollectdAmqpInstances to match the AMQP address prefix scheme. For more information, see Section 4.5.1, “Planning AMQP address prefixes”.


Remove enable-stf.yaml and ceilometer-write-qdr.yaml environment file references from your overcloud deployment. This configuration is redundant and results in duplicate information being sent from each cloud node.

  1. Create the stf-connectors.yaml file and modify it to match the AMQP address that you want for this cloud deployment:
    OS::TripleO::Services::Collectd: /usr/share/openstack-tripleo-heat-templates/deployment/metrics/collectd-container-puppet.yaml
    OS::TripleO::Services::MetricsQdr: /usr/share/openstack-tripleo-heat-templates/deployment/metrics/qdr-container-puppet.yaml
    OS::TripleO::Services::CeilometerAgentCentral: /usr/share/openstack-tripleo-heat-templates/deployment/ceilometer/ceilometer-agent-central-container-puppet.yaml
    OS::TripleO::Services::CeilometerAgentNotification: /usr/share/openstack-tripleo-heat-templates/deployment/ceilometer/ceilometer-agent-notification-container-puppet.yaml
    OS::TripleO::Services::CeilometerAgentIpmi: /usr/share/openstack-tripleo-heat-templates/deployment/ceilometer/ceilometer-agent-ipmi-container-puppet.yaml
    OS::TripleO::Services::ComputeCeilometerAgent: /usr/share/openstack-tripleo-heat-templates/deployment/ceilometer/ceilometer-agent-compute-container-puppet.yaml
    OS::TripleO::Services::Redis: /usr/share/openstack-tripleo-heat-templates/deployment/database/redis-pacemaker-puppet.yaml

    EnableSTF: true

    EventPipelinePublishers: []
    CeilometerEnablePanko: false
    CeilometerQdrPublishEvents: true
        driver: amqp
        topic: cloud1-event   1

    CollectdConnectionType: amqp1
    CollectdAmqpInterval: 5
    CollectdDefaultPollingInterval: 5

        cloud1-notify:        2
            notify: true
            format: JSON
            presettle: false
        cloud1-telemetry:     3
            format: JSON
            presettle: true

        - prefix: collectd
          distribution: multicast
        - prefix: anycast/ceilometer
          distribution: multicast

        - name: sslProfile

        - host:   4
          port: 443
          role: edge
          verifyHostname: false
          sslProfile: sslProfile

+ <1> Define the topic for Ceilometer events. This value is the address format of anycast/ceilometer/cloud1-event.sample. <2> Define the topic for collectd events. This value is the format of collectd/cloud1-notify. <3> Define the topic for collectd metrics. This value is the format of collectd/cloud1-telemetry. <4> Adjust the MetricsQdrConnectors host to the address of the STF route.

  1. Ensure that the naming convention in the stf-connectors.yaml file aligns with the spec.amqpUrl field in the Smart Gateway configuration. For example, configure the CeilometerQdrEventsConfig.topic field to a value of cloud1-event.
  2. Save the file in a directory for custom environment files, for example /home/stack/custom_templates/.
  3. Source the authentication file:

    [stack@undercloud-0 ~]$ source stackrc
    (undercloud) [stack@undercloud-0 ~]$
  4. Include the stf-connectors.yaml file in the overcloud deployment command, along with any other environment files relevant to your environment:

    (undercloud) [stack@undercloud-0 ~]$ openstack overcloud deploy \
    --templates /usr/share/openstack-tripleo-heat-templates \
    -e /home/stack/custom_templates/stf-connectors.yaml \
Additional resources

For information about validating the deployment, see Section 3.3.3, “Validating client-side installation”.

4.5.4. Querying metrics data from multiple clouds

Data stored in Prometheus has a service label attached according to the Smart Gateway it was scraped from. You can use this label to query data from a specific cloud.

To query data from a specific cloud, use a Prometheus promql query that matches the associated service label; for example: collectd_uptime{service="stf-default-collectd-telemetry-cloud1-smartgateway"}.

4.6. Ephemeral storage

Use ephemeral storage to run Service Telemetry Framework (STF) without persistently storing data in your Red Hat OpenShift Container Platform (OCP) cluster. Ephemeral storage is not recommended in a production environment due to the volatility of the data in the platform when operating correctly and as designed. For example, restarting a pod or rescheduling the workload to another node results in the loss of any local data written since the pod started.

If you enable ephemeral storage in STF, the Service Telemetry Operator does not add the relevant storage sections to the data storage components manifests.

4.6.1. Configuring ephemeral storage

To configure STF for ephemeral storage, add storageEphemeralEnabled: true to the ServiceTelemetry object in OCP. You can add storageEphemeralEnabled: true at installation time or, if you already deployed STF, complete the following steps:


  1. Log in to Red Hat OpenShift Container Platform.
  2. Change to the service-telemetry namespace:

    oc project service-telemetry
  3. Edit the ServiceTelemetry object:

    $ oc edit ServiceTelemetry stf-default
  4. Add the storageEphemeralEnabled: true parameter to the spec section:

      eventsEnabled: true
      metricsEnabled: true
      storageEphemeralEnabled: true
  5. Save your changes and close the object.

Appendix A. collectd plug-ins

This section contains a complete list of collectd plug-ins and configurations for Red Hat OpenStack Platform 16.0.

  • collectd::plugin::aggregation::aggregators
  • collectd::plugin::aggregation::interval




  • collectd::plugin::battery::values_percentage
  • collectd::plugin::battery::report_degraded
  • collectd::plugin::battery::query_state_fs
  • collectd::plugin::battery::interval
  • collectd::plugin::ceph::daemons
  • collectd::plugin::ceph::longrunavglatency
  • collectd::plugin::ceph::convertspecialmetrictypes
  • collectd::plugin::cgroups::ignore_selected
  • collectd::plugin::cgroups::interval
  • None
  • collectd::plugin::contextswitch::interval
  • collectd::plugin::cpu::reportbystate
  • collectd::plugin::cpu::reportbycpu
  • collectd::plugin::cpu::valuespercentage
  • collectd::plugin::cpu::reportnumcpu
  • collectd::plugin::cpu::reportgueststate
  • collectd::plugin::cpu::subtractgueststate
  • collectd::plugin::cpu::interval
  • None


  • collectd::plugin::csv::datadir
  • collectd::plugin::csv::storerates
  • collectd::plugin::csv::interval
  • collectd::plugin::df::devices
  • collectd::plugin::df::fstypes
  • collectd::plugin::df::ignoreselected
  • collectd::plugin::df::mountpoints
  • collectd::plugin::df::reportbydevice
  • collectd::plugin::df::reportinodes
  • collectd::plugin::df::reportreserved
  • collectd::plugin::df::valuesabsolute
  • collectd::plugin::df::valuespercentage
  • collectd::plugin::df::interval
  • collectd::plugin::disk::disks
  • collectd::plugin::disk::ignoreselected
  • collectd::plugin::disk::udevnameattr
  • collectd::plugin::disk::interval
  • collectd::plugin::entropy::interval
  • collectd::plugin::ethstat::interfaces
  • collectd::plugin::ethstat::maps
  • collectd::plugin::ethstat::mappedonly
  • collectd::plugin::ethstat::interval
  • collectd::plugin::exec::commands
  • collectd::plugin::exec::commands_defaults
  • collectd::plugin::exec::globals
  • collectd::plugin::exec::interval
  • collectd::plugin::fhcount::valuesabsolute
  • collectd::plugin::fhcount::valuespercentage
  • collectd::plugin::fhcount::interval
  • collectd::plugin::filecount::directories
  • collectd::plugin::filecount::interval
  • None
  • collectd::plugin::hddtemp::host
  • collectd::plugin::hddtemp::port
  • collectd::plugin::hddtemp::interval
  • collectd::plugin::hugepages::report_per_node_hp
  • collectd::plugin::hugepages::report_root_hp
  • collectd::plugin::hugepages::values_pages
  • collectd::plugin::hugepages::values_bytes
  • collectd::plugin::hugepages::values_percentage
  • collectd::plugin::hugepages::interval


  • collectd::plugin::interface::interfaces
  • collectd::plugin::interface::ignoreselected
  • collectd::plugin::interface::reportinactive
  • Collectd::plugin::interface::interval
  • None
  • collectd::plugin::ipmi::ignore_selected
  • collectd::plugin::ipmi::notify_sensor_add
  • collectd::plugin::ipmi::notify_sensor_remove
  • collectd::plugin::ipmi::notify_sensor_not_present
  • collectd::plugin::ipmi::sensors
  • collectd::plugin::ipmi::interval
  • collectd::plugin::irq::irqs
  • collectd::plugin::irq::ignoreselected
  • collectd::plugin::irq::interval
  • collectd::plugin::load::report_relative
  • collectd::plugin::load::interval
  • collectd::plugin::logfile::log_level
  • collectd::plugin::logfile::log_file
  • collectd::plugin::logfile::log_timestamp
  • collectd::plugin::logfile::print_severity
  • collectd::plugin::logfile::interval




  • collectd::plugin::memcached::instances
  • collectd::plugin::memcached::interval
  • collectd::plugin::memory::valuesabsolute
  • collectd::plugin::memory::valuespercentage
  • collectd::plugin::memory::interval collectd-multimeter


  • collectd::plugin::mysql::interval
  • collectd::plugin::netlink::interfaces
  • collectd::plugin::netlink::verboseinterfaces
  • collectd::plugin::netlink::qdiscs
  • collectd::plugin::netlink::classes
  • collectd::plugin::netlink::filters
  • collectd::plugin::netlink::ignoreselected
  • collectd::plugin::netlink::interval
  • collectd::plugin::network::timetolive
  • collectd::plugin::network::maxpacketsize
  • collectd::plugin::network::forward
  • collectd::plugin::network::reportstats
  • collectd::plugin::network::listeners
  • collectd::plugin::network::servers
  • collectd::plugin::network::interval
  • collectd::plugin::nfs::interval
  • collectd::plugin::ntpd::host
  • collectd::plugin::ntpd::port
  • collectd::plugin::ntpd::reverselookups
  • collectd::plugin::ntpd::includeunitid
  • collectd::plugin::ntpd::interval
  • None


  • collectd::plugin::openvpn::statusfile
  • collectd::plugin::openvpn::improvednamingschema
  • collectd::plugin::openvpn::collectcompression
  • collectd::plugin::openvpn::collectindividualusers
  • collectd::plugin::openvpn::collectusercount
  • collectd::plugin::openvpn::interval
  • collectd::plugin::ovs_events::address
  • collectd::plugin::ovs_events::dispatch
  • collectd::plugin::ovs_events::interfaces
  • collectd::plugin::ovs_events::send_notification
  • collectd::plugin::ovs_events::$port
  • collectd::plugin::ovs_events::socket
  • collectd::plugin::ovs_stats::address
  • collectd::plugin::ovs_stats::bridges
  • collectd::plugin::ovs_stats::port
  • collectd::plugin::ovs_stats::socket
  • collectd::plugin::ping::hosts
  • collectd::plugin::ping::timeout
  • collectd::plugin::ping::ttl
  • collectd::plugin::ping::source_address
  • collectd::plugin::ping::device
  • collectd::plugin::ping::max_missed
  • collectd::plugin::ping::size
  • collectd::plugin::ping::interval
  • collectd::plugin::powerdns::interval
  • collectd::plugin::powerdns::servers
  • collectd::plugin::powerdns::recursors
  • collectd::plugin::powerdns::local_socket
  • collectd::plugin::powerdns::interval
  • collectd::plugin::processes::processes
  • collectd::plugin::processes::process_matches
  • collectd::plugin::processes::collect_context_switch
  • collectd::plugin::processes::collect_file_descriptor
  • collectd::plugin::processes::collect_memory_maps
  • collectd::plugin::powerdns::interval
  • collectd::plugin::protocols::ignoreselected
  • collectd::plugin::protocols::values



  • collectd::plugin::smart::disks
  • collectd::plugin::smart::ignoreselected
  • collectd::plugin::smart::interval


  • collectd::plugin::statsd::host
  • collectd::plugin::statsd::port
  • collectd::plugin::statsd::deletecounters
  • collectd::plugin::statsd::deletetimers
  • collectd::plugin::statsd::deletegauges
  • collectd::plugin::statsd::deletesets
  • collectd::plugin::statsd::countersum
  • collectd::plugin::statsd::timerpercentile
  • collectd::plugin::statsd::timerlower
  • collectd::plugin::statsd::timerupper
  • collectd::plugin::statsd::timersum
  • collectd::plugin::statsd::timercount
  • collectd::plugin::statsd::interval
  • collectd::plugin::swap::reportbydevice
  • collectd::plugin::swap::reportbytes
  • collectd::plugin::swap::valuesabsolute
  • collectd::plugin::swap::valuespercentage
  • collectd::plugin::swap::reportio
  • collectd::plugin::swap::interval
  • collectd::plugin::syslog::log_level
  • collectd::plugin::syslog::notify_level
  • collectd::plugin::syslog::interval
  • collectd::plugin::table::tables
  • collectd::plugin::table::interval
  • collectd::plugin::tail::files
  • collectd::plugin::tail::interval
  • collectd::plugin::tail_csv::metrics
  • collectd::plugin::tail_csv::files
  • collectd::plugin::tcpconns::localports
  • collectd::plugin::tcpconns::remoteports
  • collectd::plugin::tcpconns::listening
  • collectd::plugin::tcpconns::allportssummary
  • collectd::plugin::tcpconns::interval


  • collectd::plugin::thermal::devices
  • collectd::plugin::thermal::ignoreselected
  • collectd::plugin::thermal::interval
  • collectd::plugin::threshold::types
  • collectd::plugin::threshold::plugins
  • collectd::plugin::threshold::hosts
  • collectd::plugin::threshold::interval
  • collectd::plugin::turbostat::core_c_states
  • collectd::plugin::turbostat::package_c_states
  • collectd::plugin::turbostat::system_management_interrupt
  • collectd::plugin::turbostat::digital_temperature_sensor
  • collectd::plugin::turbostat::tcc_activation_temp
  • collectd::plugin::turbostat::running_average_power_limit
  • collectd::plugin::turbostat::logical_core_names


  • collectd::plugin::uptime::interval
  • collectd::plugin::users::interval
  • collectd::plugin::uuid::uuid_file
  • collectd::plugin::uuid::interval
  • collectd::plugin::virt::connection
  • collectd::plugin::virt::refresh_interval
  • collectd::plugin::virt::domain
  • collectd::plugin::virt::block_device
  • collectd::plugin::virt::interface_device
  • collectd::plugin::virt::ignore_selected
  • collectd::plugin::virt::hostname_format
  • collectd::plugin::virt::interface_format
  • collectd::plugin::virt::extra_stats
  • collectd::plugin::virt::interval
  • collectd::plugin::vmem::verbose
  • collectd::plugin::vmem::interval



  • collectd::plugin::write_graphite::carbons
  • collectd::plugin::write_graphite::carbon_defaults
  • collectd::plugin::write_graphite::globals
  • collectd::plugin::write_kafka::kafka_host
  • collectd::plugin::write_kafka::kafka_port
  • collectd::plugin::write_kafka::kafka_hosts
  • collectd::plugin::write_kafka::topics
  • collectd::plugin::write_log::format
  • None

Legal Notice

Copyright © 2021 Red Hat, Inc.
The text of and illustrations in this document are licensed by Red Hat under a Creative Commons Attribution–Share Alike 3.0 Unported license ("CC-BY-SA"). An explanation of CC-BY-SA is available at In accordance with CC-BY-SA, if you distribute this document or an adaptation of it, you must provide the URL for the original version.
Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section 4d of CC-BY-SA to the fullest extent permitted by applicable law.
Red Hat, Red Hat Enterprise Linux, the Shadowman logo, the Red Hat logo, JBoss, OpenShift, Fedora, the Infinity logo, and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries.
Linux® is the registered trademark of Linus Torvalds in the United States and other countries.
Java® is a registered trademark of Oracle and/or its affiliates.
XFS® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States and/or other countries.
MySQL® is a registered trademark of MySQL AB in the United States, the European Union and other countries.
Node.js® is an official trademark of Joyent. Red Hat is not formally related to or endorsed by the official Joyent Node.js open source or commercial project.
The OpenStack® Word Mark and OpenStack logo are either registered trademarks/service marks or trademarks/service marks of the OpenStack Foundation, in the United States and other countries and are used with the OpenStack Foundation's permission. We are not affiliated with, endorsed or sponsored by the OpenStack Foundation, or the OpenStack community.
All other trademarks are the property of their respective owners.