Nvidia SMI and Databricks Integration

Input and output integration overview

The Nvidia SMI Plugin enables the retrieval of detailed statistics about NVIDIA GPUs attached to the host system, providing essential insights for performance monitoring.

Use Telegraf’s HTTP output plugin to push metrics straight into a Databricks Lakehouse by calling the SQL Statement Execution API with a JSON-wrapped INSERT or volume PUT command.

Integration details

Nvidia SMI

The Nvidia SMI Plugin is designed to gather metrics regarding the performance and status of NVIDIA GPUs on the host machine. By leveraging the capabilities of the nvidia-smi command-line tool, this plugin pulls crucial information such as GPU memory utilization, temperature, fan speed, and various performance metrics. This data is essential for monitoring GPU health and performance in real-time, particularly in environments where GPU performance directly impacts computing tasks, such as machine learning, 3D rendering, and high-performance computing. The plugin provides flexibility by allowing users to specify the path to the nvidia-smi binary and configure polling timeouts, accommodating both Linux and Windows systems where the nvidia-smi tool is commonly located. With its ability to collect detailed statistics on each GPU, this plugin becomes a vital resource for any infrastructure relying on NVIDIA hardware, facilitating proactive management and performance tuning.

Databricks

This configuration turns Telegraf into a lightweight ingestion agent for the Databricks Lakehouse. It leverages the Databricks SQL Statement Execution API 2.0, which accepts authenticated POST requests containing a JSON payload with a statement field. Each Telegraf flush dynamically renders a SQL INSERT (or, for file-based workflows, a PUT ... INTO /Volumes/... command) that lands the metrics into a Unity Catalog table or volume governed by Lakehouse security. Under the hood Databricks stores successful inserts as Delta Lake transactions, enabling ACID guarantees, time-travel, and scalable analytics. Operators can point the warehouse_id at any serverless or classic SQL warehouse, and all authentication is handled with a PAT or service-principal token—no agents or JDBC drivers required. Because Telegraf’s HTTP output supports custom headers, batching, TLS, and proxy settings, the same pattern scales from edge IoT gateways to container sidecars, consolidating infrastructure telemetry, application logs, or business KPIs directly into the Lakehouse for BI, ML, and Lakehouse Monitoring. Unity Catalog volumes provide a governed staging layer when file uploads and COPY INTO are preferred, and the approach aligns with Databricks’ recommended ingestion practices for partners and ISVs.

Configuration

Nvidia SMI

[[inputs.nvidia_smi]]
  ## Optional: path to nvidia-smi binary, defaults "/usr/bin/nvidia-smi"
  ## We will first try to locate the nvidia-smi binary with the explicitly specified value (or default value),
  ## if it is not found, we will try to locate it on PATH(exec.LookPath), if it is still not found, an error will be returned
  # bin_path = "/usr/bin/nvidia-smi"

  ## Optional: timeout for GPU polling
  # timeout = "5s"

Databricks

[[outputs.http]]
  ## Databricks SQL Statement Execution API endpoint
  url = "https://{{ env "DATABRICKS_HOST" }}/api/2.0/sql/statements"

  ## Use POST to submit each Telegraf batch as a SQL request
  method = "POST"

  ## Personal-access token (PAT) for workspace or service principal
  headers = { Authorization = "Bearer {{ env "DATABRICKS_TOKEN" }}" }

  ## Send JSON that wraps the metrics batch in a SQL INSERT (or PUT into a Volume)
  content_type = "application/json"

  ## Serialize metrics as JSON so they can be embedded in the SQL statement
  data_format = "json"
  json_timestamp_units = "1ms"

  ## Build the request body.  Telegraf replaces the template variables at runtime.
  ## Example inserts a row per metric into a Unity-Catalog table.
  body_template = """
  {
    \"statement\": \"INSERT INTO ${TARGET_TABLE} VALUES {{range .Metrics}}(from_unixtime({{.timestamp}}/1000), {{.fields.usage}}, '{{.tags.host}}'){{end}}\",
    \"warehouse_id\": \"${WAREHOUSE_ID}\"
  }
  """

  ## Optional: add batching limits or TLS settings
  # batch_size = 500
  # timeout     = "10s"

Input and output integration examples

Nvidia SMI

1. Real-Time GPU Monitoring for ML Training: Continuously monitor the GPU utilization and memory usage during machine learning model training. This enables data scientists to ensure that their GPUs are not being overutilized or underutilized, optimizing resource allocation and reviewing performance bottlenecks in real-time.

2. Automated Alerts for Overheating GPUs: Implement a system using the Nvidia SMI plugin to track GPU temperatures and set alerts for instances where temperatures exceed safe thresholds. This proactive monitoring can prevent hardware damage and improve system reliability by alerting administrators to potential cooling issues before they result in failure.

3. Performance Baselines for GPU Resources: Establish baseline performance metrics for your GPU resources. By regularly collecting data and analyzing trends in GPU usage, organizations can identify anomalies and optimize their workloads accordingly, leading to enhanced operational efficiency.

4. Dockerized GPU Usage Insights: In a containerized environment, use the plugin to monitor GPU performance from within a Docker container. This allows developers to track GPU performance of their applications in production, facilitating troubleshooting and performance optimization within isolated environments.

Databricks

1. Edge-to-Lakehouse Telemetry Pipe: Deploy Telegraf on factory PLCs to sample vibration metrics and post them every second to a serverless SQL warehouse. Delta tables power PowerBI dashboards that alert engineers when thresholds drift.
2. Blue-Green CI/CD Rollout Metrics: Attach a Telegraf sidecar to each Kubernetes canary pod; it inserts container stats into a Unity Catalog table tagged by deployment_id, letting Databricks SQL compare error-rate percentiles and auto-rollback underperforming versions.
3. SaaS Usage Metering: Insert per-tenant API-call counters via the HTTP plugin; a nightly Lakehouse query aggregates usage into invoices, eliminating custom metering micro-services.
4. Security Forensics Lake: Upload JSON batches of Suricata IDS events to a Unity Catalog volume using PUT commands, then run COPY INTO for near-real-time enrichment with Delta Live Tables, producing a searchable threat-intel lake that joins network logs with user session data.

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.