Storage Adapters

Telegen’s storage adapters collect metrics from enterprise storage arrays via vendor APIs.

Overview

Storage adapters support:

Vendor	Products	API
Dell	PowerStore, PowerScale	REST API
HPE	Primera, 3PAR	WSAPI
Pure Storage	FlashArray, FlashBlade	REST API v2
NetApp	ONTAP, E-Series	ONTAP REST API

Dell PowerStore

Configuration

collector:
  storage:
    dell:
      enabled: true
      poll_interval: 60s
      
      targets:
        - name: "powerstore-prod-01"
          address: "https://powerstore.example.com"
          username: "monitor"
          password: "${DELL_PASSWORD}"
          verify_ssl: true
          
          metrics:
            performance: true
            capacity: true
            alerts: true
            hardware: true

Metrics Collected

Metric	Description
`dell_powerstore_volume_read_iops`	Read IOPS per volume
`dell_powerstore_volume_write_iops`	Write IOPS per volume
`dell_powerstore_volume_read_latency_us`	Read latency (microseconds)
`dell_powerstore_volume_write_latency_us`	Write latency (microseconds)
`dell_powerstore_volume_size_bytes`	Volume total size
`dell_powerstore_volume_used_bytes`	Volume used space
`dell_powerstore_cluster_capacity_total_bytes`	Total cluster capacity
`dell_powerstore_cluster_capacity_used_bytes`	Used cluster capacity
`dell_powerstore_data_reduction_ratio`	Data reduction ratio
`dell_powerstore_hardware_health`	Hardware component health

Labels

Label	Description
`array`	Array name
`volume`	Volume name
`appliance`	Appliance ID
`host`	Host attachment

HPE Primera/3PAR

Configuration

collector:
  storage:
    hpe:
      enabled: true
      poll_interval: 60s
      
      targets:
        - name: "primera-prod"
          address: "https://primera.example.com:8080"
          username: "monitor"
          password: "${HPE_PASSWORD}"
          verify_ssl: true
          
          metrics:
            system: true
            cpgs: true
            volumes: true
            hosts: true
            ports: true

Metrics Collected

Metric	Description
`hpe_primera_capacity_total_bytes`	Total system capacity
`hpe_primera_capacity_allocated_bytes`	Allocated capacity
`hpe_primera_capacity_free_bytes`	Free capacity
`hpe_primera_efficiency_ratio`	Compaction ratio
`hpe_primera_volume_size_bytes`	Volume size
`hpe_primera_volume_used_bytes`	Volume used space
`hpe_primera_cpg_capacity_bytes`	CPG capacity
`hpe_primera_port_bandwidth_bytes`	Port bandwidth
`hpe_primera_node_cpu_percent`	Node CPU usage
`hpe_primera_node_cache_hit_percent`	Cache hit ratio

Labels

Label	Description
`array`	Array name
`volume`	Volume name
`cpg`	Common provisioning group
`node`	Controller node
`port`	FC/iSCSI port

Pure Storage FlashArray

Configuration

collector:
  storage:
    pure:
      enabled: true
      poll_interval: 60s
      
      targets:
        - name: "pure-prod-01"
          address: "https://purestorage.example.com"
          api_token: "${PURE_API_TOKEN}"
          
          metrics:
            arrays: true
            volumes: true
            hosts: true
            pods: true
            replication: true

Generating API Token

# On Pure Storage array
pureuser@array01> pureapitoken create --user monitor

Metrics Collected

Metric	Description
`pure_flasharray_capacity_bytes`	Total array capacity
`pure_flasharray_used_bytes`	Used capacity
`pure_flasharray_data_reduction`	Data reduction ratio
`pure_flasharray_volume_size_bytes`	Volume provisioned size
`pure_flasharray_volume_used_bytes`	Volume used space
`pure_flasharray_volume_read_iops`	Volume read IOPS
`pure_flasharray_volume_write_iops`	Volume write IOPS
`pure_flasharray_volume_read_latency_us`	Read latency
`pure_flasharray_volume_write_latency_us`	Write latency
`pure_flasharray_pod_status`	Pod replication status
`pure_flasharray_hardware_status`	Hardware health

Labels

Label	Description
`array`	Array name
`volume`	Volume name
`host`	Connected host
`pod`	Replication pod
`drive`	Drive bay

NetApp ONTAP (Harvest-Parity Collector)

Telegen includes a Harvest-parity ONTAP collector that provides comprehensive observability for NetApp ONTAP storage arrays. This collector is a drop-in replacement for NetApp Harvest with full metric catalog coverage.

Collector Types

Collector	Engine	Description
Rest	ONTAP REST API	65+ inventory/configuration objects (capacity, state, config)
RestPerf	Counter Tables API	54+ performance counter tables with rate cooking
KeyPerf	Resource Endpoints	Performance via `statistics.*` (Volume always; fallback for ASA r2, GCNV)
EMS	Event Management System	EMS events → OTLP logs (86 message types)

Auto-Selection Logic

Telegen automatically probes ONTAP capabilities and selects the appropriate collectors:

RESTPerf enabled when api/cluster/counter/tables responds (ONTAP ~9.11.1+)
RESTPerf disabled for ASA r2 (disaggregated), GCNV mode, or older clusters → KeyPerf takes over
Default collectors: [rest, restperf, keyperf, ems]

ONTAP Object Coverage

Rest Catalog (65 Objects)

Category	Objects
Core	Cluster, Node, SVM, Volume, Aggregate, LUN, Namespace, Qtree, Snapshot, Quota
Network	LIF, NetPort, NetRoute, EthernetSwitchPort, NetConnections, NIC
Protocols	CIFS Session/Share, NFS Clients, iSCSI, FCP, OntapS3, OntapS3Policy
Data Protection	SnapMirror, SnapMirrorPolicy, SnapshotPolicy
Security	Security, SecurityAccount, SecurityCert, SecurityLogin, SecuritySsh, SecurityAuditDestination
Hardware	Disk, Shelf, Sensor, FRU, License
Management	ExportRule, Lock, MAVRequest, Mediator, NtpServer, ClusterSchedule, ClusterSoftware
v3 Opt-In	AuditLog, CIFSSession, CIFSShare, ExportRule, NFSClients, NDMPSession, NetConnections

RestPerf Catalog (54+ Objects)

Protocol	Objects
NFS	NFSv3, NFSv4, NFSv4.1, NFSv4.2 (per-node and per-SVM), NFSv4Pool
SMB/CIFS	CIFSNode, CIFSvserver, SMB2
Block	iSCSI, FCP, FCPPort, FCPLif, FCVI
NVMe	Namespace, NVMfLif, NVMfRdmaPort, NVMfTcpPort
Object	OntapS3SVM
System	SystemNode, HostAdapter, Disk, Path, NicCommon, Netstat
WAFL	WAFL, WAFLAggr, WAFLSizer, WAFLCompBin
QoS	Workload, WorkloadVolume
Other	TokenManager, CopyManager, FlexCachePerf, Vscan, VscanSVM, FPolicy, Netstat

KeyPerf Catalog (14 Objects)

Volume, Aggregate, CIFSvserver, Cluster, EthernetSwitchPort, LUN, Namespace, NFSv3, NFSv4, NFSv4.1, SystemNode, Qtree (opt-in)

Configuration

telegen:
  mode: collector  # or unified

netapp_ontap:
  - name: "ontap-prod-01"
    address: "https://10.0.10.140"
    username: "monitor"
    password: "${NETAPP_PASSWORD}"
    verify_ssl: true
    timeout: 30s
    coverage: full                    # full | harvest_default
    collectors: [rest, restperf, keyperf, ems]
    templates_dir: ""                 # empty = configs/netapp
    ems:
      enabled: true
      resolve_after: 672h
    labels:
      environment: "production"

# NetApp E-Series (SANtricity)
netapp_eseries:
  - name: "eseries-prod-01"
    address: "https://10.0.10.150"
    username: "monitor"
    password: "${NETAPP_ESERIES_PASSWORD}"
    verify_ssl: true
    timeout: 30s
    labels:
      environment: "production"

Required Permissions

Create a read-only monitoring role:

# On ONTAP cluster
security login role create -role monitor -cmddirname "volume show" -access readonly
security login role create -role monitor -cmddirname "aggregate show" -access readonly
security login role create -role monitor -cmddirname "lun show" -access readonly
security login role create -role monitor -cmddirname "statistics" -access readonly
security login create -user-or-group-name monitor -role monitor -application http -authmethod password

Performance Metrics (Exhaustive)

Cluster

Metric	Description
`cluster_read_ops`, `cluster_write_ops`, `cluster_total_ops`	IOPS
`cluster_read_data`, `cluster_write_data`, `cluster_total_data`	Throughput
`cluster_read_latency`, `cluster_write_latency`, `cluster_total_latency`	Latency
`cluster_space_available`	Available space

Volume

Metric	Description
`volume_size`, `volume_size_used`, `volume_size_used_percent`	Capacity
`volume_read_ops`, `volume_write_ops`, `volume_total_ops`	IOPS
`volume_read_data`, `volume_write_data`, `volume_total_data`	Throughput
`volume_read_latency`, `volume_write_latency`, `volume_total_latency`	Latency
`volume_nfs_total_ops`, `volume_sis_*_saved_percent`	NFS ops, dedupe savings

Node

Metric	Description
`node_avg_processor_busy`, `node_cpu_busy`	CPU
`node_total_ops`, `node_read_ops`, `node_write_ops`	IOPS
`node_net_data_recv`, `node_net_data_sent`	Network
`node_disk_data_read`, `node_disk_data_written`	Disk

LUN / Namespace

Metric	Description
`lun_size`, `lun_size_used`	Capacity
`lun_read_ops`, `lun_write_ops`, `lun_total_ops`	IOPS
`lun_read_data`, `lun_write_data`	Throughput
`namespace_size`, `namespace_read_ops`	NVMe namespace metrics

Plugin System

Telegen includes a plugin system for computed metrics:

Plugin	Description
MetricAgent	`compute_metric` expressions (ADD, SUBTRACT, MULTIPLY, DIVIDE, PERCENT)
LabelAgent	Convert label strings to numeric metrics
Volume	Marks `style` label; derives NFS ops totals
Aggregator	Rolls up metrics by label (e.g., node-level summaries)
Health	Ensures synthetic `status` metric
Sensor	Hardware sensor metrics (temperature, fan, power)
Shelf	Shelf hardware inventory
Nic	NIC utilization
Fcp	FCP utilization
FabricPool	Cloud tier metrics
VolumeTopClients	Top clients/files/users
VolumeAnalytics	Activity analytics
QoS	Workload throughput bounds

EMS Events (OTLP Logs)

NetApp EMS events are exported as OTLP log records:

log_record:
  body: "Volume vol01 is nearly full (95% used)"
  severity: WARNING
  attributes:
    netapp.ems.message_name: "wafl.vol.full"
    netapp.ems.severity: "WARNING"
    netapp.cluster.name: "ontap-prod-01"
    netapp.volume.name: "vol01"
    telegen.signal.category: "Storage Logs"
    telegen.signal.subcategory: "NetApp EMS"

E-Series Support (SANtricity)

NetApp E-Series arrays are supported via the SANtricity REST API:

Collector	Objects
E-Series Rest	Volume, Array, Host, Hardware, SsdCache, Pool, Workload (7 objects)
E-Series Perf	Volume, Controller, Drive, Pool, Array, SsdCache, Interface, Application, Workload (9 objects)

Dashboards

Six pre-built dashboards are available in docs/reference/netapp-ontap-dashboards.md:

Fleet Overview (9 panels) — Arrays, nodes, aggregates, volumes, capacity
Node Performance (8 panels) — CPU, IOPS, latency, throughput, network, disk
Aggregate & Volume Capacity (7 panels) — Capacity, free, used %, physical used
Volume / LUN Performance (8 panels) — IOPS, throughput, latency, top volumes
Data Protection / SnapMirror (4 panels) — Lag time, last transfer, failed updates
EMS / Events (7 panels) — Live EMS stream, severity, top messages, hardware events

Metric Parity

The NetApp ONTAP collector targets 1,558 unique Harvest-compatible metric families covering all Rest/RestPerf/KeyPerf templates, aggregator plugins, computed metrics, and ZAPI legacy aliases.

Labels

Label	Description
`cluster`	Cluster name
`node`	Node name
`aggregate`	Aggregate name
`volume`	Volume name
`svm`	Storage VM name
`lun`	LUN path
`array`	Array name (E-Series)

Common Dashboards

Capacity Planning

# Total capacity across all arrays
sum(storage_capacity_total_bytes) by (vendor)

# Capacity utilization
sum(storage_capacity_used_bytes) / sum(storage_capacity_total_bytes) * 100

# Days until full (at current growth rate)
(storage_capacity_total_bytes - storage_capacity_used_bytes)
/ deriv(storage_capacity_used_bytes[7d])
/ 86400

Performance

# Total IOPS across arrays
sum(rate(storage_volume_read_iops[5m]) + rate(storage_volume_write_iops[5m]))

# Average latency
avg(storage_volume_read_latency_us + storage_volume_write_latency_us) / 2

# Top 10 volumes by IOPS
topk(10, rate(storage_volume_read_iops[5m]) + rate(storage_volume_write_iops[5m]))

Health Alerts

groups:
  - name: storage
    rules:
      - alert: StorageArrayCapacityHigh
        expr: storage_capacity_used_bytes / storage_capacity_total_bytes > 0.85
        for: 5m
        labels:
          severity: warning
        annotations:
          summary: "Storage array {{ $labels.array }} is over 85% capacity"
      
      - alert: StorageVolumeLatencyHigh
        expr: storage_volume_read_latency_us > 10000
        for: 5m
        labels:
          severity: warning
        annotations:
          summary: "Volume {{ $labels.volume }} has high latency"
      
      - alert: StorageHardwareFailure
        expr: storage_hardware_status != 1
        for: 1m
        labels:
          severity: critical
        annotations:
          summary: "Hardware issue on {{ $labels.array }}"

Security Considerations

Credential Management

Use environment variables or secrets:

collector:
  storage:
    pure:
      targets:
        - name: "pure-prod"
          address: "https://pure.example.com"
          # Reference environment variable
          api_token: "${PURE_API_TOKEN}"

In Kubernetes:

apiVersion: v1
kind: Secret
metadata:
  name: storage-credentials
type: Opaque
stringData:
  PURE_API_TOKEN: "your-token-here"
  DELL_PASSWORD: "your-password"
  NETAPP_PASSWORD: "your-password"

Network Security

Use HTTPS with valid certificates
Restrict collector IP access on storage arrays
Use read-only monitoring accounts
Rotate credentials regularly

Multi-Array Example

telegen:
  mode: collector
  service_name: "storage-collector"

otlp:
  endpoint: "otel-collector:4317"

collector:
  storage:
    # Dell PowerStore
    dell:
      enabled: true
      poll_interval: 60s
      targets:
        - name: "powerstore-dc1"
          address: "https://10.0.10.100"
          username: "monitor"
          password: "${DELL_PASSWORD}"
        - name: "powerstore-dc2"
          address: "https://10.0.20.100"
          username: "monitor"
          password: "${DELL_PASSWORD}"
    
    # Pure Storage
    pure:
      enabled: true
      poll_interval: 60s
      targets:
        - name: "pure-prod"
          address: "https://10.0.10.110"
          api_token: "${PURE_TOKEN_PROD}"
        - name: "pure-dev"
          address: "https://10.0.10.111"
          api_token: "${PURE_TOKEN_DEV}"
    
    # NetApp
    netapp:
      enabled: true
      poll_interval: 60s
      targets:
        - name: "ontap-nas"
          address: "https://10.0.10.120"
          username: "monitor"
          password: "${NETAPP_PASSWORD}"

Next Steps

SNMP Receiver - Network device monitoring
Collector Mode Configuration - Collector configuration
Monitoring Telegen - Storage monitoring dashboards