Add OLM CSV metrics

docs/data-collection.md

@@ -105,7 +105,13 @@ For the OpenShift 4 Developer Preview we will be sending back these exact attrib
   // subscription_sync_total is the number of times an OLM operator
   // Subscription has been synced, labelled by name and installed csv
   '{__name__="subscription_sync_total"}',
-  //
+  // csv_succeeded is unique to the namespace, name, version, and phase labels.
+  // The metrics is always present and can be equal to 0 or 1, where 0 represents that the
+  // csv is not in the succeeded state while 1 represents that the csv is in the succeeded state.
+  '{__name__="csv_succeeded"}',
+  // csv_abnormal represents the reason why a csv is not in the succeeded state and includes the
+  // namespace, name, version, phase, reason labels. When a csv is updated, the previous time series associated with the csv will be deleted.
+  '{__name__="csv_abnormal"}',
   // OCS metrics to be collected:
   // ceph_cluster_total_bytes gives the size of ceph cluster in bytes.
   '{__name__="ceph_cluster_total_bytes"}',

docs/sample-metrics.md

@@ -13,7 +13,7 @@ return the full set of metrics that the Telemeter client captures:
 
 [embedmd]:# (telemeter_query txt)
 ```txt
-{__name__="up"} or {__name__="cluster_version"} or {__name__="cluster_version_available_updates"} or {__name__="cluster_operator_up"} or {__name__="cluster_operator_conditions"} or {__name__="cluster_version_payload"} or {__name__="cluster_installer"} or {__name__="cluster_infrastructure_provider"} or {__name__="cluster_feature_set"} or {__name__="node_uname_info"} or {__name__="instance:etcd_object_counts:sum"} or {__name__="ALERTS",alertstate="firing"} or {__name__="code:apiserver_request_count:rate:sum"} or {__name__="cluster:capacity_cpu_cores:sum"} or {__name__="cluster:capacity_memory_bytes:sum"} or {__name__="cluster:cpu_usage_cores:sum"} or {__name__="cluster:memory_usage_bytes:sum"} or {__name__="openshift:cpu_usage_cores:sum"} or {__name__="openshift:memory_usage_bytes:sum"} or {__name__="workload:cpu_usage_cores:sum"} or {__name__="workload:memory_usage_bytes:sum"} or {__name__="cluster:virt_platform_nodes:sum"} or {__name__="cluster:node_instance_type_count:sum"} or {__name__="cnv:vmi_status_running:count"} or {__name__="node_role_os_version_machine:cpu_capacity_cores:sum"} or {__name__="node_role_os_version_machine:cpu_capacity_sockets:sum"} or {__name__="subscription_sync_total"} or {__name__="ceph_cluster_total_bytes"} or {__name__="ceph_cluster_total_used_raw_bytes"} or {__name__="ceph_health_status"} or {__name__="job:ceph_osd_metadata:count"} or {__name__="job:kube_pv:count"} or {__name__="job:ceph_pools_iops:total"} or {__name__="job:ceph_pools_iops_bytes:total"} or {__name__="job:ceph_versions_running:count"} or {__name__="job:noobaa_total_unhealthy_buckets:sum"} or {__name__="job:noobaa_bucket_count:sum"} or {__name__="job:noobaa_total_object_count:sum"} or {__name__="noobaa_accounts_num"} or {__name__="noobaa_total_usage"} or {__name__="console_url"} or {__name__="cluster:network_attachment_definition_instances:max"} or {__name__="cluster:network_attachment_definition_enabled_instance_up:max"}
+{__name__="up"} or {__name__="cluster_version"} or {__name__="cluster_version_available_updates"} or {__name__="cluster_operator_up"} or {__name__="cluster_operator_conditions"} or {__name__="cluster_version_payload"} or {__name__="cluster_installer"} or {__name__="cluster_infrastructure_provider"} or {__name__="cluster_feature_set"} or {__name__="node_uname_info"} or {__name__="instance:etcd_object_counts:sum"} or {__name__="ALERTS",alertstate="firing"} or {__name__="code:apiserver_request_count:rate:sum"} or {__name__="cluster:capacity_cpu_cores:sum"} or {__name__="cluster:capacity_memory_bytes:sum"} or {__name__="cluster:cpu_usage_cores:sum"} or {__name__="cluster:memory_usage_bytes:sum"} or {__name__="openshift:cpu_usage_cores:sum"} or {__name__="openshift:memory_usage_bytes:sum"} or {__name__="workload:cpu_usage_cores:sum"} or {__name__="workload:memory_usage_bytes:sum"} or {__name__="cluster:virt_platform_nodes:sum"} or {__name__="cluster:node_instance_type_count:sum"} or {__name__="cnv:vmi_status_running:count"} or {__name__="node_role_os_version_machine:cpu_capacity_cores:sum"} or {__name__="node_role_os_version_machine:cpu_capacity_sockets:sum"} or {__name__="subscription_sync_total"} or {__name__="csv_succeeded"} or {__name__="csv_abnormal"} or {__name__="ceph_cluster_total_bytes"} or {__name__="ceph_cluster_total_used_raw_bytes"} or {__name__="ceph_health_status"} or {__name__="job:ceph_osd_metadata:count"} or {__name__="job:kube_pv:count"} or {__name__="job:ceph_pools_iops:total"} or {__name__="job:ceph_pools_iops_bytes:total"} or {__name__="job:ceph_versions_running:count"} or {__name__="job:noobaa_total_unhealthy_buckets:sum"} or {__name__="job:noobaa_bucket_count:sum"} or {__name__="job:noobaa_total_object_count:sum"} or {__name__="noobaa_accounts_num"} or {__name__="noobaa_total_usage"} or {__name__="console_url"} or {__name__="cluster:network_attachment_definition_instances:max"} or {__name__="cluster:network_attachment_definition_enabled_instance_up:max"}
 ```
 
 For reference, here is an example response produced by a running OpenShift cluster:

jsonnet/telemeter/metrics.jsonnet

@@ -97,7 +97,13 @@
   // subscription_sync_total is the number of times an OLM operator
   // Subscription has been synced, labelled by name and installed csv
   '{__name__="subscription_sync_total"}',
-  //
+  // csv_succeeded is unique to the namespace, name, version, and phase labels.
+  // The metrics is always present and can be equal to 0 or 1, where 0 represents that the
+  // csv is not in the succeeded state while 1 represents that the csv is in the succeeded state.
+  '{__name__="csv_succeeded"}',
+  // csv_abnormal represents the reason why a csv is not in the succeeded state and includes the
+  // namespace, name, version, phase, reason labels. When a csv is updated, the previous time series associated with the csv will be deleted.
+  '{__name__="csv_abnormal"}',
   // OCS metrics to be collected:
   // ceph_cluster_total_bytes gives the size of ceph cluster in bytes.
   '{__name__="ceph_cluster_total_bytes"}',

manifests/benchmark/statefulSetTelemeterServer.yaml

@@ -52,6 +52,8 @@ spec:
         - --whitelist={__name__="node_role_os_version_machine:cpu_capacity_cores:sum"}
         - --whitelist={__name__="node_role_os_version_machine:cpu_capacity_sockets:sum"}
         - --whitelist={__name__="subscription_sync_total"}
+        - --whitelist={__name__="csv_succeeded"}
+        - --whitelist={__name__="csv_abnormal"}
         - --whitelist={__name__="ceph_cluster_total_bytes"}
         - --whitelist={__name__="ceph_cluster_total_used_raw_bytes"}
         - --whitelist={__name__="ceph_health_status"}

manifests/client/deployment.yaml

@@ -54,6 +54,8 @@ spec:
         - --match={__name__="node_role_os_version_machine:cpu_capacity_cores:sum"}
         - --match={__name__="node_role_os_version_machine:cpu_capacity_sockets:sum"}
         - --match={__name__="subscription_sync_total"}
+        - --match={__name__="csv_succeeded"}
+        - --match={__name__="csv_abnormal"}
         - --match={__name__="ceph_cluster_total_bytes"}
         - --match={__name__="ceph_cluster_total_used_raw_bytes"}
         - --match={__name__="ceph_health_status"}