Configure the OTLP receiver to collect ChromaDB metrics

Configure the Splunk Distribution of the OpenTelemetry Collector to send ChromaDB telemetry data to Splunk Observability Cloud.

You can monitor the performance of your ChromaDB vector databases by configuring the Splunk Distribution of the OpenTelemetry Collector to send ChromaDB telemetry data to Splunk Observability Cloud.

ChromaDB includes a built-in OpenTelemetry integration module that enables the collection and export of metrics. This integration sends telemetry data to the OTLP receiver, which gets data in to the Splunk Distribution of the OpenTelemetry Collector.

To use this data integration, you must have a running ChromaDB server.
  1. Deploy the Splunk Distribution of the OpenTelemetry Collector to your host or container platform:
  2. Configure the following environment variables on the ChromaDB server to enable OpenTelemetry export:
    CODE 
    CHROMA_OPEN_TELEMETRY__ENDPOINT="<OTLP_ENDPOINT>" 
CHROMA_OPEN_TELEMETRY__SERVICE_NAME="<SERVICE_NAME>"
  3. Restart the ChromaDB service to apply the configuration.

Configuration settings

Configuration settings for the OTLP receiver.

To view the configuration settings for the OTLP receiver, see Settings.

Metrics

The following metrics are available for ChromaDB.

These metrics are considered custom metrics in Splunk Observability Cloud.

Metric name Type Description
add_retries counter The number of retry attempts made when adding items to a collection.
cache_hit counter The number of successful lookups served from cache.
cache_miss counter The number of lookups that were not found in cache.
collection_add counter The number of add operations performed on a collection.
collection_delete counter The number of delete operations performed on a collection.
collection_get counter The number of get/retrieval operations performed on a collection.
collection_update counter The number of update operations performed on a collection.
component_queue_depth histogram The depth of the component's message queue.
create_collection counter The number of collection creation operations performed.
delete_collection counter The number of collection deletion operations performed.
delete_retries counter The number of retry attempts made when deleting items.
get_collection counter The number of operations requesting access to a collection.
get_database counter The number of operations requesting metadata or state for the database.
get_latency histogram The latency distribution for get operations.
get_retries counter The number of retry attempts made for get operations.
get_tenant counter The number of tenant lookup operations performed.
get_user_identity counter The number of user identity lookup operations performed.
insert_latency histogram The latency distribution for insert operations.
metering_events_sent.read counter The number of read metering events sent by the frontend to the metering event receiver.
metering_events_sent.write counter The number of write metering events sent by the frontend to the metering event receiver.
pre_flight_checks counter The number of pre-flight validation or readiness checks performed.
update_retries counter The number of retry attempts made when updating items.

Attributes

The following resource attributes for ChromaDB are derived from the metric source.

Attribute name Description
host.name The hostname of the machine where ChromaDB is running.
service.name The logical service identifier in OpenTelemetry. Set by OTEL_SERVICE_NAME.
component_name Name of the subsystem component inside ChromaDB whose task queue is being measured.