OpenTelemetry

Send metrics and heartbeat events from Akto's hybrid modules to an OpenTelemetry endpoint.

Prerequisites

Before setting up the OpenTelemetry Integration, ensure you have the HTTP header name and value (API key) for authentication with the OpenTelemetry endpoint.

Quick Setup Steps

1

Access Integrations

• Go to Settings > Integrations.

• Find and click "Configure" next to OpenTelemetry.

  

2

Enter OpenTelemetry (OTLP) Endpoint Details

1. Enter OTLP endpoint.

2. Enter authentication details.

   1. Enter Header name.

   2. Enter API Key.

   If your OpenTelemetry endpoint requires a prefix before the API Key, please include the prefix when entering the API Key.

   For example: Basic YOUR_API_KEY_HERE

   

3

Save Configuration

• Click "Save" to finalise.

Data Reference

Akto sends two types of telemetry to your OpenTelemetry endpoint: metrics (flushed every 120 seconds) and heartbeat events (sent every 30 seconds per module).

All signals are emitted with service.name = akto-telemetry on the OTLP resource.

Heartbeat Event

Each Akto module emits an OTLP log record with event.name set to AktoModuleHeartbeatReceived every 30 seconds, sent to the /v1/logs endpoint. In your OTLP backend, filter log records where event.name = AktoModuleHeartbeatReceived to query them.

The log record carries the following attributes:

Attribute	Description
akto.account.id	Akto account ID
akto.account.name	Akto account name
akto.organization.id	Akto organization ID
akto.organization.name	Akto organization name
akto.module.id	Unique ID of the module instance
akto.module.type	Module type (e.g. MINI_RUNTIME, MINI_TESTING)
akto.module.name	Human-readable module name
akto.module.currentVersion	Deployed version of the module
akto.module.startedTs	Unix timestamp when the module started
akto.module.lastHeartbeatReceived	Unix timestamp of the last heartbeat

Metrics

All metrics are prefixed akto.metric. and flushed every 120 seconds to the /v1/metrics endpoint. Each metric includes the same module and account attributes listed in the heartbeat table above. Metrics are emitted as either a LongCounter or DoubleGauge OTLP instrument.

Runtime metrics

Emitted by the Mini-Runtime modules.

Metric name	OTLP instrument	Description
akto.metric.rt_kafka_record_count	LongCounter	Number of records processed by the runtime module
akto.metric.rt_kafka_record_size	LongCounter	Total byte size of records processed by the runtime module
akto.metric.rt_kafka_latency	DoubleGauge	Average processing latency for runtime records
akto.metric.rt_api_received_count	LongCounter	Number of APIs received by the mini-runtime module
akto.metric.kafka_records_lag_max	DoubleGauge	Maximum consumer lag across Kafka partitions
akto.metric.kafka_records_consumed_rate	DoubleGauge	Current rate of Kafka record consumption
akto.metric.kafka_fetch_avg_latency	DoubleGauge	Average fetch latency from Kafka
akto.metric.kafka_bytes_consumed_rate	DoubleGauge	Current byte consumption rate from Kafka
akto.metric.cyborg_new_api_count	LongCounter	Newly discovered APIs detected in the flush window
akto.metric.cyborg_total_api_count	LongCounter	Total APIs tracked by Cyborg
akto.metric.delta_catalog_new_count	LongCounter	New items added to the delta catalog
akto.metric.delta_catalog_total_count	LongCounter	Total items in the delta catalog
akto.metric.cyborg_api_payload_size	LongCounter	Total size of API payloads processed by Cyborg

Testing metrics

Emitted by the Mini-Testing module.

Metric name	OTLP instrument	Description
akto.metric.testing_run_count	LongCounter	Test runs executed in the flush window
akto.metric.testing_run_latency	DoubleGauge	Average time to complete a test run
akto.metric.sample_data_fetch_latency	DoubleGauge	Average latency for single sample-data fetches
akto.metric.multiple_sample_data_fetch_latency	DoubleGauge	Average latency for bulk sample-data fetches

Cyborg metrics

Emitted by any module that calls Cyborg.

Metric name	OTLP instrument	Description
akto.metric.cyborg_call_count	LongCounter	Calls made to Cyborg
akto.metric.cyborg_call_latency	DoubleGauge	Average Cyborg call latency
akto.metric.cyborg_data_size	LongCounter	Total data size sent to/from Cyborg

Infrastructure metrics

Emitted by every module, reflecting the JVM process running that module.

Metric name	OTLP instrument	Description
akto.metric.cpu_usage_percent	DoubleGauge	Process CPU usage (%)
akto.metric.heap_memory_used_mb	DoubleGauge	JVM heap memory in use (MB)
akto.metric.heap_memory_max_mb	DoubleGauge	Maximum JVM heap memory available (MB)
akto.metric.non_heap_memory_used_mb	DoubleGauge	JVM non-heap memory in use (MB)
akto.metric.thread_count	DoubleGauge	Number of live JVM threads
akto.metric.available_processors	DoubleGauge	CPU cores available to the JVM
akto.metric.total_physical_memory_mb	DoubleGauge	Total physical host memory (MB)

Get Support for your Akto setup

There are multiple ways to request support from Akto. We are 24X7 available on the following:

1. In-app intercom support. Message us with your query on intercom in Akto dashboard and someone will reply.

2. Join our discord channel for community support.

3. Contact [e-mail] for email support.

4. Contact us here.