Fluentd and OpenObserve Integration

Input and output integration overview

The Fluentd Input Plugin gathers metrics from Fluentd’s in_monitor plugin endpoint. It provides insights into various plugin metrics while allowing for custom configurations to reduce series cardinality.

This configuration pairs Telegraf’s HTTP output with OpenObserve’s native JSON ingestion API, turning any Telegraf agent into a first-class OpenObserve collector.

Integration details

Fluentd

This plugin gathers metrics from the Fluentd plugin endpoint provided by the in_monitor plugin. It reads data from the /api/plugin.json resource and allows exclusion of specific plugins based on their type.

OpenObserve

OpenObserve is an open source observability platform written in Rust that stores data cost-effectively on object storage or local disk. It exposes REST endpoints such as /api/{org}/ingest/metrics/_json that accept batched metric documents conforming to a concise JSON schema, making it an attractive drop-in replacement for Loki or Elasticsearch stacks. The Telegraf HTTP output plugin streams metrics to arbitrary HTTP targets; when the "data_format = "json"" serializer is selected, Telegraf batches its metric objects into a payload that matches OpenObserve’s ingestion contract. The plugin supports configurable batch size, custom headers, TLS, and compression, allowing operators to authenticate with Basic or Bearer tokens and to enforce back-pressure without additional collectors. By reusing existing Telegraf agents already collecting system, application, or SNMP data, organizations can funnel rich telemetry into OpenObserve dashboards and SQL-like analytics with minimal overhead, enabling unified observability, long-term retention, and real-time alerting without vendor lock-in.

Configuration

Fluentd

[[inputs.fluentd]]
  ## This plugin reads information exposed by fluentd (using /api/plugins.json endpoint).
  ##
  ## Endpoint:
  ## - only one URI is allowed
  ## - https is not supported
  endpoint = "http://localhost:24220/api/plugins.json"

  ## Define which plugins have to be excluded (based on "type" field - e.g. monitor_agent)
  exclude = [
    "monitor_agent",
    "dummy",
  ]

OpenObserve

[[outputs.http]]
  ## OpenObserve JSON metrics ingestion endpoint
  url = "https://api.openobserve.ai/api/default/ingest/metrics/_json"

  ## Use POST to push batches
  method = "POST"

  ## Basic auth header (base64 encoded "username:password")
  headers = { Authorization = "Basic dXNlcjpwYXNzd29yZA==" }

  ## Timeout for HTTP requests
  timeout = "10s"

  ## Override Content-Type to match OpenObserve expectation
  content_type = "application/json"

  ## Force Telegraf to batch and serialize metrics as JSON
  data_format = "json"

  ## JSON serializer specific options
  json_timestamp_units = "1ms"

  ## Uncomment to restrict batch size
  # batch_size = 5000

Input and output integration examples

Fluentd

1. Basic Configuration: Set up the Fluentd Input Plugin to gather metrics from your Fluentd instance’s monitoring endpoint, ensuring you are able to track performance and usage statistics.
2. Excluding Plugins: Use the exclude option to ignore specific plugins’ metrics that are not necessary for your monitoring needs, streamlining data collection and focusing on what matters.
3. Custom Plugin ID: Implement the @id parameter in your Fluentd configuration to maintain a consistent plugin_id, which helps avoid issues with high series cardinality during frequent restarts.

OpenObserve

1. Edge Device Health Mirror: Deploy Telegraf on thousands of industrial IoT devices to capture temperature, vibration, and power metrics, then use this output to push JSON batches to OpenObserve. Plant operators gain a real-time overview of machine health and can trigger maintenance based on anomalies without relying on heavyweight collectors.

2. Blue-Green Deployment Canary: Attach a lightweight Telegraf sidecar to each Kubernetes release-candidate pod that scrapes /metrics and forwards container stats to a dedicated “canary” stream in OpenObserve. Continuous comparison of error rates between blue and green versions empowers the CI pipeline to auto-roll back poor performers within seconds.

3. Multi-Tenant SaaS Billing Pipeline: Emit per-customer usage counters via Telegraf and tag them with tenant_id; the HTTP plugin posts them to OpenObserve where SQL reports aggregate usage into invoices, eliminating separate metering services and simplifying compliance audits.

4. Security Threat Scoring: Fuse Suricata events and host resource metrics in Telegraf, deliver them to OpenObserve’s analytics engine, and run stream-processing rules that correlate spikes in suspicious traffic with CPU saturation to produce an actionable threat score and automatically open tickets in a SOAR platform.

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.