Hashicorp Vault and OSI PI Integration

Input and output integration overview

The Hashicorp Vault plugin for Telegraf allows for the collection of metrics from Hashicorp Vault services, facilitating monitoring and operational insights.

This setup converts Telegraf into a lightweight PI Web API publisher, letting you push any Telegraf metric into the OSI PI System with a simple HTTP POST.

Integration details

Hashicorp Vault

The Hashicorp Vault plugin is designed to collect metrics from Vault agents running within a cluster. It enables Telegraf, an agent for collecting and reporting metrics, to interface with the Vault services, typically listening on a local address such as http://127.0.0.1:8200. This plugin requires a valid token for authorization, ensuring secure access to the Vault API. Users must configure either a token directly or provide a path to a token file, enhancing flexibility in authentication methods. Proper configuration of the timeout and optional TLS settings further relates to the security and responsiveness of the metrics collection process. As Vault is a critical tool in managing secrets and protecting sensitive data, monitoring its performance and health through this plugin is essential for maintaining operational security and efficiency.

OSI PI

OSI PI is an data management and analytics platform used in energy, manufacturing, and critical infrastructure. The PI Web API is its REST interface, exposing endpoints such as /piwebapi/streams/{WebId}/value that accept JSON payloads containing a Timestamp and Value. By pairing Telegraf’s flexible HTTP output with this endpoint, any metric Telegraf collects—SNMP counters, Modbus readings, Kubernetes stats—can be written directly into PI without installing proprietary interfaces. The configuration above authenticates with Basic or Kerberos, serializes each batch to JSON, and renders a minimal body template that aligns with PI Web API’s single-value write contract. Because Telegraf already supports batching, TLS, proxies, and custom headers, this approach scales from edge gateways to cloud VMs, allowing organizations to back-fill historical data, stream live telemetry, or mirror non-PI sources (e.g., Prometheus) into the PI data archive. It also sidesteps older SDK dependencies and enables hybrid architectures where PI remains on-prem while Telegraf agents run in containers or IIoT devices.

Configuration

Hashicorp Vault

[[inputs.vault]]
  ## URL for the Vault agent
  # url = "http://127.0.0.1:8200"

  ## Use Vault token for authorization.
  ## Vault token configuration is mandatory.
  ## If both are empty or both are set, an error is thrown.
  # token_file = "/path/to/auth/token"
  ## OR
  token = "s.CDDrgg5zPv5ssI0Z2P4qxJj2"

  ## Set response_timeout (default 5 seconds)
  # response_timeout = "5s"

  ## Optional TLS Config
  # tls_ca = /path/to/cafile
  # tls_cert = /path/to/certfile
  # tls_key = /path/to/keyfile

OSI PI

[[outputs.http]]
  ## PI Web API endpoint for writing a single value to a PI Point by Web ID
  url = "https://${PI_HOST}/piwebapi/streams/${WEB_ID}/value"

  ## Use POST for each batch
  method       = "POST"
  content_type = "application/json"

  ## Basic-auth header (base64-encoded "DOMAIN\\user:password")
  headers = { Authorization = "Basic ${BASIC_AUTH}" }

  ## Serialize Telegraf metrics as JSON
  data_format           = "json"
  json_timestamp_units  = "1ms"

  ## Render the JSON body that PI Web API expects
  body_template = """
  {{ range .Metrics -}}
  { "Timestamp": "{{ .timestamp | formatDate \"2006-01-02T15:04:05Z07:00\" }}", "Value": {{ index .fields 0 }} }
  {{ end -}}
  """

  ## Tune networking / batching if needed
  # timeout     = "10s"
  # batch_size  = 1

Input and output integration examples

Hashicorp Vault

1. Centralized Secret Management Monitoring: Utilize the Vault plugin to monitor multiple Vault instances across a distributed system, allowing for a unified view of secret access patterns and system health. This setup can help DevOps teams quickly identify any anomalies in secret access, providing essential insights into security postures across different environments.

2. Audit Logging Integration: Configure this plugin to feed monitoring metrics into an audit logging system, enabling organizations to have a comprehensive view of their Vault interactions. By correlating audit logs with metrics, teams can investigate issues, optimize performance, and ensure compliance with security policies more effectively.

3. Performance Benchmarking During Deployments: During application deployments that interact with Vault, use the plugin to monitor the effects of those deployments on Vault performance. This allows engineering teams to understand how changes impact secret management workflows and to proactively address performance bottlenecks, ensuring smooth deployment processes.

4. Alerting for Threshold Exceedance: Integrate this plugin with alerting mechanisms to notify administrators when metrics exceed predefined thresholds. This proactive monitoring can help teams respond swiftly to potential issues, maintaining system reliability and uptime by allowing them to take action before any serious incidents arise.

OSI PI

1. Remote Pump Stations Telemetry Bridge: Install Telegraf on edge gateways at oil-field pump stations, gather flow-meter and vibration readings over Modbus, and POST them to the PI Web API. Operations teams view real-time data in PI Vision without deploying heavyweight PI interfaces, while bandwidth-friendly batching keeps satellite links economical.

2. Green-Energy Micro-Grid Dashboard: Export inverter, battery, and weather metrics from MQTT into Telegraf, which relays them to PI. PI AF analytics can calculate real-time power balance and feed a campus dashboard; historical deltas inform sustainability reports.

3. Brownfield SCADA Modernization: Legacy PLCs logged to CSV are ingested by Telegraf’s tail input; each row is parsed and immediately sent to PI via HTTP, creating a live data stream that co-exists with archival files while the SCADA upgrade proceeds incrementally.

4. Synthetic Data Generator for Training: Telegraf’s exec input can run a script that emits simulated sensor patterns. Posting those metrics to a non-production PI server through the Web API supplies realistic datasets for PI Vision training sessions without risking production tags.

Feedback

Thank you for being part of our community! If you have any general feedback or found any bugs on these pages, we welcome and encourage your input. Please submit your feedback in the InfluxDB community Slack.