OpenTelemetry and Apache Hudi Integration

Input and output integration overview

This plugin receives traces, metrics, and logs from OpenTelemetry clients and agents via gRPC, enabling comprehensive observability of applications.

Writes metrics to Parquet files via Telegraf’s Parquet output plugin, preparing them for ingestion into Apache Hudi’s lakehouse architecture.

Integration details

OpenTelemetry

The OpenTelemetry plugin is designed to receive telemetry data such as traces, metrics, and logs from clients and agents implementing OpenTelemetry via gRPC. This plugin initiates a gRPC service that listens for incoming telemetry data, making it distinct from standard plugins that collect metrics at defined intervals. The OpenTelemetry ecosystem aids developers in observing and understanding their applications’ performance by providing a vendor-neutral way to instrument, generate, collect, and export telemetry data. Key features of this plugin include customizable connection timeouts, adjustable maximum message sizes for incoming data, and options for specifying span, log, and profile dimensions to tag the incoming metrics. With this flexibility, organizations can tailor their telemetry collection to meet precise observability requirements and ensure seamless data integration into systems like InfluxDB.

Apache Hudi

This configuration leverages Telegraf’s Parquet plugin to serialize metrics into columnar Parquet files suitable for downstream ingestion by Apache Hudi. The plugin writes metrics grouped by metric name into files in a specified directory, buffering writes for efficiency and optionally rotating files on timers. It considers schema compatibility—metrics with incompatible schemas are dropped—ensuring consistency. Apache Hudi can then consume these Parquet files via tools like DeltaStreamer or Spark jobs, enabling transactional ingestion, time-travel queries, and upserts on your time series data.

Configuration

OpenTelemetry

[[inputs.opentelemetry]]
  ## Override the default (0.0.0.0:4317) destination OpenTelemetry gRPC service
  ## address:port
  # service_address = "0.0.0.0:4317"

  ## Override the default (5s) new connection timeout
  # timeout = "5s"

  ## gRPC Maximum Message Size
  # max_msg_size = "4MB"

  ## Override the default span attributes to be used as line protocol tags.
  ## These are always included as tags:
  ## - trace ID
  ## - span ID
  ## Common attributes can be found here:
  ## - https://github.com/open-telemetry/opentelemetry-collector/tree/main/semconv
  # span_dimensions = ["service.name", "span.name"]

  ## Override the default log record attributes to be used as line protocol tags.
  ## These are always included as tags, if available:
  ## - trace ID
  ## - span ID
  ## Common attributes can be found here:
  ## - https://github.com/open-telemetry/opentelemetry-collector/tree/main/semconv
  ## When using InfluxDB for both logs and traces, be certain that log_record_dimensions
  ## matches the span_dimensions value.
  # log_record_dimensions = ["service.name"]

  ## Override the default profile attributes to be used as line protocol tags.
  ## These are always included as tags, if available:
  ## - profile_id
  ## - address
  ## - sample
  ## - sample_name
  ## - sample_unit
  ## - sample_type
  ## - sample_type_unit
  ## Common attributes can be found here:
  ## - https://github.com/open-telemetry/opentelemetry-collector/tree/main/semconv
  # profile_dimensions = []

  ## Override the default (prometheus-v1) metrics schema.
  ## Supports: "prometheus-v1", "prometheus-v2"
  ## For more information about the alternatives, read the Prometheus input
  ## plugin notes.
  # metrics_schema = "prometheus-v1"

  ## Optional TLS Config.
  ## For advanced options: https://github.com/influxdata/telegraf/blob/v1.18.3/docs/TLS.md
  ##
  ## Set one or more allowed client CA certificate file names to
  ## enable mutually authenticated TLS connections.
  # tls_allowed_cacerts = ["/etc/telegraf/clientca.pem"]
  ## Add service certificate and key.
  # tls_cert = "/etc/telegraf/cert.pem"
  # tls_key = "/etc/telegraf/key.pem"

Apache Hudi

[[outputs.parquet]]
  ## Directory to write parquet files in. If a file already exists the output
  ## will attempt to continue using the existing file.
  directory = "/var/lib/telegraf/hudi_metrics"

  ## File rotation interval (default is no rotation)
  # rotation_interval = "1h"

  ## Buffer size before writing (default is 1000 metrics)
  # buffer_size = 1000

  ## Optional: compression codec (snappy, gzip, etc.)
  # compression_codec = "snappy"

  ## When grouping metrics, each metric name goes to its own file
  ## If a metric’s schema doesn’t match the existing schema, it will be dropped

Input and output integration examples

OpenTelemetry

1. Unified Monitoring Across Services: Use the OpenTelemetry plugin to collect and consolidate telemetry data from various microservices within a Kubernetes environment. By instrumenting each service with OpenTelemetry, you can utilize this plugin to gather a holistic view of application performance and dependencies in real-time, enabling faster troubleshooting and improved reliability of complex systems.

2. Enhanced Debugging with Traces: Implement this plugin to capture end-to-end traces of requests flowing through multiple services. For instance, when a user initiates a transaction that triggers several backend services, the OpenTelemetry plugin can record detailed traces that highlight performance bottlenecks, giving developers the necessary insights to debug issues and optimize their code.

3. Dynamic Load Testing and Performance Monitoring: Leverage the capabilities of this plugin during load testing phases by collecting live metrics and traces under simulated higher loads. This approach helps to evaluate the resilience of the application components and identify potential performance degradations preemptively, ensuring a smooth user experience in production.

4. Integrated Logging and Metrics for Real-Time Monitoring: Combine the OpenTelemetry plugin with logging frameworks to gather real-time logs alongside metric data, creating a powerful observability platform. For example, integrate it within a CI/CD pipeline to monitor builds and deployments, while collecting logs that help diagnose failures or performance issues in real-time.

Apache Hudi

1. Transactional Lakehouse Metrics: Buffer and write Web service metrics as Parquet files for DeltaStreamer to ingest into Hudi, enabling upserts, ACID compliance, and time-travel on historical performance data.

2. Edge Device Batch Analytics: Telegraf running on IoT gateways writes metrics to Parquet locally, where periodic Spark jobs ingest them into Hudi for long-term analytics and traceability.

3. Schema-Enforced Abnormal Metric Handling: Use Parquet plugin’s strict schema-dropping behavior to prevent malformed or unexpected metric changes. Hudi ingestion then guarantees consistent schema and data quality in downstream datasets.

4. Data Platform Integration: Store Telegraf metrics as Parquet files in an S3/ADLS landing zone. Hudi’s Spark-based ingestion pipeline then loads them into a unified, queryable lakehouse with business events and logs.

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.