Documentation Grafana Cloud Instrument and send data Grafana Alloy Reference Components otelcol otelcol.exporter.kafka

otelcol.exporter.kafka

otelcol.exporter.kafka accepts logs, metrics, and traces telemetry data from other otelcol components and sends it to Kafka.

It is important to use otelcol.exporter.kafka together with otelcol.processor.batch to make sure otelcol.exporter.kafka doesn’t slow down due to sending Kafka a huge number of small payloads.

Multiple otelcol.exporter.kafka components can be specified by giving them different labels.

Usage

alloy Copy

otelcol.exporter.kafka "LABEL" {
  protocol_version = "PROTOCOL_VERSION"
}

Arguments

The following arguments are supported:

If topic is not set, different topics will be used for different telemetry signals:

• Metrics will be sent to an otlp_metrics topic.
• Traces will be sent to an otlp_spans topic.
• Logs will be sent to an otlp_logs topic.

If topic is set, the same topic will be used for all telemetry signals - metrics, logs, and traces.

When topic_from_attribute is set, it will take precedence over topic.

The encoding argument determines how to encode messages sent to Kafka. encoding must be one of the following strings:

• Encodings which work for traces, logs, and metrics:
  • "otlp_proto": Encode messages as OTLP protobuf.
  • "otlp_json": Encode messages as OTLP JSON.
• Encodings which work only for traces:
  • "jaeger_proto": The payload is serialized to a single Jaeger proto Span, and keyed by TraceID.
  • "jaeger_json": The payload is serialized to a single Jaeger JSON Span using jsonpb, and keyed by TraceID.
  • "zipkin_proto": The payload is serialized to Zipkin v2 proto Span.
  • "zipkin_json": The payload is serialized to Zipkin v2 JSON Span.
• Encodings which work only for logs:
  • "raw": If the log record body is a byte array, it is sent as is. Otherwise, it is serialized to JSON. Resource and record attributes are discarded.

partition_traces_by_id does not have any effect on Jaeger encoding exporters since Jaeger exporters include trace ID as the message key by default.

Blocks

The following blocks are supported inside the definition of otelcol.exporter.kafka:

The > symbol indicates deeper levels of nesting. For example, authentication > tls refers to a tls block defined inside an authentication block.

authentication block

The authentication block holds the definition of different authentication mechanisms to use when connecting to Kafka brokers. It doesn’t support any arguments and is configured fully through inner blocks.

plaintext block

The plaintext block configures plain text authentication against Kafka brokers.

The following arguments are supported:

sasl block

The sasl block configures SASL authentication against Kafka brokers.

The following arguments are supported:

The mechanism argument can be set to one of the following strings:

• "PLAIN"
• "AWS_MSK_IAM"
• "SCRAM-SHA-256"
• "SCRAM-SHA-512"

When mechanism is set to "AWS_MSK_IAM", the aws_msk child block must also be provided.

You can set the version argument to either 0 or 1.

aws_msk block

The aws_msk block configures extra parameters for SASL authentication when using the AWS_MSK_IAM mechanism.

The following arguments are supported:

tls block

The tls block configures TLS settings used for connecting to the Kafka brokers. If the tls block isn’t provided, TLS won’t be used for communication.

The following arguments are supported:

If the server doesn’t support TLS, you must set the insecure argument to true.

To disable tls for connections to the server, set the insecure argument to true.

If reload_interval is set to "0s", the certificate never reloaded.

The following pairs of arguments are mutually exclusive and can’t both be set simultaneously:

• ca_pem and ca_file
• cert_pem and cert_file
• key_pem and key_file

If cipher_suites is left blank, a safe default list is used. Refer to the Go TLS documentation for a list of supported cipher suites.

kerberos block

The kerberos block configures Kerberos authentication against the Kafka broker.

The following arguments are supported:

When use_keytab is false, the password argument is required. When use_keytab is true, the file pointed to by the keytab_file argument is used for authentication instead. At most one of password or keytab_file must be provided.

disable_fast_negotiation is useful for Kerberos implementations which don’t support PA-FX-FAST (Pre-Authentication Framework - Fast) negotiation.

metadata block

The metadata block configures how to retrieve and store metadata from the Kafka broker.

The following arguments are supported:

If the include_all_topics argument is true, a full set of metadata for all topics is maintained rather than the minimal set that has been necessary so far. Including the full set of metadata is more convenient for users but can consume a substantial amount of memory if you have many topics and partitions.

Retrieving metadata may fail if the Kafka broker is starting up at the same time as the Alloy component. The retry child block can be provided to customize retry behavior.

retry block

The retry block configures how to retry retrieving metadata when retrieval fails.

The following arguments are supported:

retry_on_failure block

The retry_on_failure block configures how failed requests to Kafka are retried.

The following arguments are supported:

When enabled is true, failed batches are retried after a given interval. The initial_interval argument specifies how long to wait before the first retry attempt. If requests continue to fail, the time to wait before retrying increases by the factor specified by the multiplier argument, which must be greater than 1.0. The max_interval argument specifies the upper bound of how long to wait between retries.

The randomization_factor argument is useful for adding jitter between retrying Alloy instances. If randomization_factor is greater than 0, the wait time before retries is multiplied by a random factor in the range [ I - randomization_factor * I, I + randomization_factor * I], where I is the current interval.

If a batch hasn’t been sent successfully, it’s discarded after the time specified by max_elapsed_time elapses. If max_elapsed_time is set to "0s", failed requests are retried forever until they succeed.

queue block

The queue block configures an in-memory buffer of batches before data is sent to the gRPC server.

The following arguments are supported:

When enabled is true, data is first written to an in-memory buffer before sending it to the configured server. Batches sent to the component’s input exported field are added to the buffer as long as the number of unsent batches doesn’t exceed the configured queue_size.

queue_size determines how long an endpoint outage is tolerated. Assuming 100 requests/second, the default queue size 1000 provides about 10 seconds of outage tolerance. To calculate the correct value for queue_size, multiply the average number of outgoing requests per second by the time in seconds that outages are tolerated. A very high value can cause Out Of Memory (OOM) kills.

The num_consumers argument controls how many readers read from the buffer and send data in parallel. Larger values of num_consumers allow data to be sent more quickly at the expense of increased network traffic.

producer block

The producer block configures how to retry retrieving metadata when retrieval fails.

The following arguments are supported:

Refer to the sarama documentation for more information on required_acks.

compression could be set to either none, gzip, snappy, lz4, or zstd. Refer to the Sarama documentation for more information.

debug_metrics block

The debug_metrics block configures the metrics that this component generates to monitor its state.

The following arguments are supported:

disable_high_cardinality_metrics is the Grafana Alloy equivalent to the telemetry.disableHighCardinalityMetrics feature gate in the OpenTelemetry Collector. It removes attributes that could cause high cardinality metrics. For example, attributes with IP addresses and port numbers in metrics about HTTP and gRPC connections are removed.

level is the Alloy equivalent to the telemetry.metrics.level feature gate in the OpenTelemetry Collector. Possible values are "none", "basic", "normal" and "detailed".

Exported fields

The following fields are exported and can be referenced by other components:

input accepts otelcol.Consumer data for any telemetry signal (metrics, logs, or traces).

Component health

otelcol.exporter.kafka is only reported as unhealthy if given an invalid configuration.

Debug information

otelcol.exporter.kafka does not expose any component-specific debug information.

Example

This example forwards telemetry data through a batch processor before finally sending it to Kafka:

alloy Copy

otelcol.receiver.otlp "default" {
  http {}
  grpc {}

  output {
    metrics = [otelcol.processor.batch.default.input]
    logs    = [otelcol.processor.batch.default.input]
    traces  = [otelcol.processor.batch.default.input]
  }
}

otelcol.processor.batch "default" {
  output {
    metrics = [otelcol.exporter.kafka.default.input]
    logs    = [otelcol.exporter.kafka.default.input]
    traces  = [otelcol.exporter.kafka.default.input]
  }
}

otelcol.exporter.kafka "default" {
  brokers          = ["localhost:9092"]
  protocol_version = "2.0.0"
}

Compatible components

otelcol.exporter.kafka has exports that can be consumed by the following components:

• Components that consume OpenTelemetry otelcol.Consumer