Skip to content

PowerScale Metrics

This page catalogs the metrics collected by the Omnia PowerScale telemetry pipeline via the CSM Observability framework. These metrics are stored in VictoriaMetrics for visualization in Grafana.

Collection method

Property Value
Collection tool CSM Metrics for PowerScale (Karavi Observability)
Protocol OneFS REST API → OpenTelemetry Collector → Prometheus scrape
Default interval 30 seconds (configurable via scrape_interval in CSM Observability values.yaml)
Storage VictoriaMetrics time-series database

Performance metrics

Metric Unit Description
powerscale_cluster_protocol_read_iops IOPS Read I/O operations per second at the protocol level (NFS, SMB, S3).
powerscale_cluster_protocol_write_iops IOPS Write I/O operations per second at the protocol level.
powerscale_cluster_protocol_read_bw Bytes/s Read throughput (bytes per second) at the protocol level.
powerscale_cluster_protocol_write_bw Bytes/s Write throughput (bytes per second) at the protocol level.
powerscale_cluster_protocol_read_latency Microseconds Average read latency per protocol operation.
powerscale_cluster_protocol_write_latency Microseconds Average write latency per protocol operation.

Capacity metrics

Metric Unit Description
powerscale_cluster_capacity_total_bytes Bytes Total raw storage capacity of the PowerScale cluster.
powerscale_cluster_capacity_used_bytes Bytes Storage capacity currently consumed.
powerscale_cluster_capacity_avail_bytes Bytes Storage capacity available for allocation.
powerscale_node_capacity_total_bytes Bytes Total capacity per node.
powerscale_node_capacity_used_bytes Bytes Capacity consumed per node.

Health metrics

Metric Unit Description
powerscale_node_status Enum Node health status (online, offline, degraded).
powerscale_disk_health Enum Physical disk health status per node.
powerscale_cluster_rebalance_status Enum Cluster data rebalance/restripe status.
powerscale_protection_group_status Enum Protection group health.

CSI health monitor metrics

When the CSI PowerScale health monitor is enabled, the following additional metrics are collected:

PV Metrics:

Metric Unit Description
powerscale_volume_status Enum PV phase (1=Bound, 0=Other). Labels: pv_name, phase.
powerscale_volume_count Count Total PowerScale PVs by phase.
powerscale_volume_capacity_bytes Bytes PV capacity in bytes.
powerscale_volume_info Info PV metadata. Labels: pv_name, phase, storage_class, reclaim_policy, access_modes, volume_handle, pvc_name, pvc_namespace.
powerscale_volume_age_seconds Seconds Seconds since PV creation.

PVC Metrics:

Metric Unit Description
powerscale_pvc_status_phase Enum PVC phase (1=Bound, 0=Other). Labels: pvc_name, pvc_namespace, phase.
powerscale_pvc_requested_bytes Bytes PVC requested storage in bytes.
powerscale_pvc_count Count Total PowerScale PVCs by phase.

Health Event Metrics:

Metric Unit Description
powerscale_volume_health_abnormal Enum Volume condition abnormal (1=abnormal, 0=healthy).
powerscale_volume_abnormal_events_total Count Total VolumeConditionAbnormal events.
powerscale_node_failure_events_total Count Total node failure events.
powerscale_node_ready Enum Node Ready condition (1=True, 0=False).

Storage Class Metrics:

Metric Unit Description
powerscale_storageclass_info Info StorageClass metadata. Labels: storageclass, provisioner, reclaim_policy, volume_binding_mode, allow_volume_expansion.
powerscale_total_capacity_bytes Bytes Total capacity of all PowerScale PVs in bytes.

Metric labels

All PowerScale metrics include the following common labels:

Label Description
cluster PowerScale cluster name.
node PowerScale node identifier.
protocol Protocol type (NFS, SMB, S3) for protocol-level metrics.

Info