Metrics Reference

Reference of the metrics that each Feast component exports:

For how to configure Feast to export Metrics, see the Metrics user guide.

Feast Core

Exported Metrics

Feast Core exports the following metrics:

Metrics

Description

Tags

feast_core_request_latency_seconds

Feast Core's latency in serving Requests in Seconds.

service, method, status_code

feast_core_feature_set_total

No. of Feature Sets registered with Feast Core.

None

feast_core_store_total

No. of Stores registered with Feast Core.

None

feast_core_max_memory_bytes

Max amount of memory the Java virtual machine will attempt to use.

None

feast_core_total_memory_bytes

Total amount of memory in the Java virtual machine

None

feast_core_free_memory_bytes

Total amount of free memory in the Java virtual machine.

None

feast_core_gc_collection_seconds

Time spent in a given JVM garbage collector in seconds.

None

Metric Tags

Exported Feast Core metrics may be filtered by the following tags/keys

Tag

Description

service

Name of the Service that request is made to. Should be set to CoreService

method

Name of the Method that the request is calling. (ie ListFeatureSets)

status_code

Status code returned as a result of handling the requests (ie OK). Can be used to find request failures.

Feast Serving

Exported Metrics

Feast Serving exports the following metrics:

Metric

Description

Tags

feast_serving_request_latency_seconds

Feast Serving's latency in serving Requests in Seconds.

method

feast_serving_request_feature_count

No. of requests retrieving a Feature from Feast Serving.

project, feature_name

feast_serving_not_found_feature_count

No. of requests retrieving a Feature has resulted in a NOT_FOUND field status.

project, feature_name

feast_serving_stale_feature_count

No. of requests retrieving a Feature resulted in a OUTSIDE_MAX_AGE field status.

project, feature_name

feast_serving_grpc_request_count

Total gRPC requests served.

method

Metric Tags

Exported Feast Serving metrics may be filtered by the following tags/keys

Tag

Description

method

Name of the Method that the request is calling. (ie ListFeatureSets)

status_code

Status code returned as a result of handling the requests (ie OK). Can be used to find request failures.

project

Name of the project that the FeatureSet of the Feature retrieved belongs to.

feature_name

Name of the Feature being retrieved.

Feast Ingestion Job

Feast Ingestion computes both metrics an statistics on data ingestion. Make sure you familar with data ingestion concepts before proceeding.

For documentation on Feature value statistics computed by the Ingestion Job see Statistics

Metrics Namespace

Metrics are computed at two stages of the Feature Row's/Feature Value's life cycle when being processed by the Ingestion Job:

  • Inflight- Prior to writing data to stores, but after successful validation of data.

  • WriteToStoreSucess- After a successful store write.

Metrics processed by each staged will be tagged with metrics_namespace to the stage where the metric was computed.

Metrics Bucketing

Metrics with a {BUCKET} are computed on a 60 second window. Suffix with the following to select the histogram bucket to use:

  • min - minimum value.

  • max - maximum value.

  • mean- mean value.

  • percentile_90- 90 percentile.

  • percentile_95- 95 percentile.

  • percentile_99- 99 percentile.

Exported Metrics

Metric

Description

Tags

feast_ingestion_feature_row_lag_ms_{BUCKET}

Lag time in milliseconds between succeeding ingested Feature Rows.

feast_store, feast_project_name,feast_featureSet_name,ingestion_job_name,

metrics_namespace

feast_ingestion_feature_value_lag_ms_{BUCKET}

Lag time in milliseconds between succeeding ingested values for each Feature.

feast_store, feast_project_name,feast_featureSet_name,

feast_feature_name,

ingestion_job_name,

metrics_namespace

feast_ingestion_feature_value_{BUCKET}

Last value feature for each Feature.

feast_store, feast_project_name,feast_featureSet_name,

feast_feature_name,

ingestion_job_name,

metrics_namespace

feast_ingestion_feature_row_ingested_count

No. of Ingested Feature Rows

feast_store, feast_project_name,feast_featureSet_name,ingestion_job_name,

metrics_namespace

feast_ingestion_feature_value_missing_count

No. of times a ingested Feature values did not provide a value for the Feature.

feast_store, feast_project_name,feast_featureSet_name,

feast_feature_name,

ingestion_job_name,

metrics_namespace

feast_ingestion_deadletter_row_count

No. of Feature Rows that that the Ingestion Job did not successfully write to store.

feast_store, feast_project_name,feast_featureSet_name,ingestion_job_name

Metric Tags

Exported Feast Ingestion Job metrics may be filtered by the following tags/keys

Tag

Description

feast_store

Name of the target store the Ingestion Job is writing to.

feast_project_name

Name of the project that the ingested FeatureSet belongs to.

feast_featureSet_name

Name of the Feature Set being ingested.

feast_feature_name

Name of the Feature being ingested.

ingestion_job_name

Name of the Ingestion Job performing data ingestion. Typically this is set to the Id of the Ingestion Job.

metrics_namespace

Stage where metrics where computed. Either Inflight or WriteToStoreSuccess