How netapp/harvest works
NetApp Harvest collects performance, capacity, and hardware metrics from ONTAP, StorageGRID, E-Series, and network devices, transforms them, and exports them to time-series backends. The principal runtime path is configured poller → collectors and plugins → metric matrices → exporters. The repository also includes local-system monitoring, operational tools, and an MCP server for querying Harvest-collected metrics and managing alert rules. Relationships below are limited to README statements and source-index references or shown excerpts; unsampled collector details are represented by their enclosing directories.
Main entry points: poller.go Poller orchestration · matrix.go Metric matrices · client.go E-Series collection · collector.go Collector runtime
Components 24
Collection runtime
- Poller configuration
pkg/conf/conf.go - Poller orchestration
cmd/poller/poller.go - Collector runtime
cmd/poller/collector/collector.go - Plugin framework
cmd/poller/plugin/plugin.go - Metric matrices
pkg/matrix/matrix.go
Target collectors
- ONTAP REST collectors
cmd/collectors/rest - ONTAP ZAPI collectors
cmd/collectors/zapi/collector/zapi.go - ONTAP performance
cmd/collectors/restperf/restperf.go - E-Series collection
cmd/collectors/eseries/rest/client.go - E-Series hardware
cmd/collectors/eseries/plugins/hardware/controller.go - E-Series performance
cmd/collectors/eseriesperf/plugins/controller/controller.go - StorageGRID collection
cmd/collectors/storagegrid/storagegrid.go - StorageGRID API client
cmd/collectors/storagegrid/rest/client.go - Network device collectors
cmd/collectors/cisco/rest/client.go - Local system metrics
cmd/collectors/unix/main.go
Metrics and outputs
- InfluxDB exporter
cmd/exporters/influxdb/influxdb.go - Prometheus exporter
cmd/exporters/prometheus/prometheus.go - VictoriaMetrics exporter
cmd/exporters/victoriametrics/victoriametrics.go
Operations and query
- Harvest command
cmd/harvest/harvest.go - REST diagnostic tool
cmd/tools/rest/client.go - MCP server
mcp/cmd/server/main.go - Alert rule manager
mcp/pkg/rules/manager.go
Other components
- Operator
- Prometheus service
Connections 32
- Operator → Poller configuration: configures
README.md - Poller configuration → Poller orchestration: configures
cmd/poller/poller.go - Poller orchestration → Collector runtime: starts collectors
cmd/poller/poller.go - Poller orchestration → ONTAP REST collectors: dispatches to
cmd/poller/poller.go - Poller orchestration → ONTAP ZAPI collectors: dispatches to
cmd/poller/poller.go - Poller orchestration → ONTAP performance: dispatches to
cmd/poller/poller.go - Poller orchestration → E-Series collection: dispatches to
cmd/poller/poller.go - Poller orchestration → E-Series performance: dispatches to
cmd/poller/poller.go - Poller orchestration → StorageGRID collection: dispatches to
cmd/poller/poller.go - Poller orchestration → Network device collectors: dispatches to
cmd/poller/poller.go - Poller orchestration → Local system metrics: dispatches to
cmd/poller/poller.go - Collector runtime → Plugin framework: uses
cmd/poller/collector/collector.go - Collector runtime → Metric matrices: records metrics
cmd/poller/collector/collector.go - ONTAP REST collectors → Metric matrices: produces metrics
- ONTAP ZAPI collectors → Metric matrices: produces metrics
- ONTAP performance → Metric matrices: produces metrics
- E-Series collection → Metric matrices: produces metrics
- E-Series collection → E-Series hardware: uses plugin
- E-Series performance → Metric matrices: transforms metrics
- E-Series performance → E-Series collection: fetches data
cmd/collectors/eseriesperf/plugins/controller/controller.go - StorageGRID collection → StorageGRID API client: requests data
cmd/collectors/storagegrid/storagegrid.go - StorageGRID collection → Metric matrices: builds metrics
cmd/collectors/storagegrid/storagegrid.go - Network device collectors → Metric matrices: produces metrics
- Local system metrics → Metric matrices: updates metrics
cmd/collectors/unix/main.go - Poller orchestration → InfluxDB exporter: routes metrics
README.md - Poller orchestration → Prometheus exporter: routes metrics
- Poller orchestration → VictoriaMetrics exporter: routes metrics
cmd/poller/poller.go - Operator → Harvest command: runs
cmd/harvest/harvest.go - Operator → REST diagnostic tool: inspects targets
- Operator → MCP server: queries metrics
README.md - MCP server → Prometheus service: queries
- MCP server → Alert rule manager: manages rules