Metrics
The metrics table contains the details of ingestions, materialized views, and continuous exports of an Eventhouse KQL database, which is part of Real-Time Intelligence. For each metric, a log event record is stored in the EventhouseMetrics table.
Metric operation logs
Use the metrics to:
- Analyze ingestion performance and trends.
- Monitor batch vs streaming ingestions.
- Troubleshoot ingestion failures.
- Deep dive into ingestion flows.
- Materialized views monitoring and health.
- Continuous exports monitoring.
The following table describes the columns stored in the EventhouseMetrics table:
| Column Name | Type | Description |
|---|---|---|
| CapacityId | string | The Fabric capacity identifier. |
| CustomerTenantId | string | The customer tenant identifier. |
| DurationMs | long | Not applicable. |
| ItemId | string | The identifier of the Fabric Eventhouse item |
| ItemKind | string | The type of the Fabric item. Valid values: Eventhouse. |
| ItemName | string | The name of the Fabric Eventhouse item. |
| Level | string | Not applicable. |
| MetricCount | long | The metric count value. |
| MetricMaxValue | long | The metric maximum value. |
| MetricMinValue | long | The metric minimum value. |
| MetricName | string | The metric name. |
| MetricSpecificDimensions | dynamic | The specific dimensions of each metric, as described in Metric Specific Dimension Column. Where relevant, dimension descriptions are provided as part of the metric description. |
| MetricSumValue | long | The metric sum value. |
| OperationName | string | The name of the operation performed. |
| Region | string | The region where the Fabric KQL database is located. |
| Timestamp | datetime | The time (UTC) the event was generated. |
| WorkspaceId | string | The identifier of the workspace. |
| WorkspaceMonitoringTableName | string | The name of the workspace monitoring table. Valid values: EventhouseQueryLogs |
| WorkspaceName | string | The name of the workspace. |
Metric Specific Dimension Column
The following table contains a list of all the reported Eventhouse metrics, and the specific dimensions reported for each metric.
| Metric Type | MetricName | Unit | Aggregation | Description | Metric Specific Dimensions |
|---|---|---|---|---|---|
| Ingestion | BatchBlobCount | Count | Avg, Max, Min | The number of data sources ingested in a completed batch. | Database, Table |
| Ingestion | BatchDurationSec | Seconds | Avg, Max, Min | The duration of the batching phase within the ingestion flow. | Database, Table |
| Ingestion | BatchSizeBytes | Bytes | Avg, Max, Min | The expected uncompressed data size in an aggregated ingestion batch. | Database, Table |
| Ingestion | BatchesProcessed | Count | Sum, Max, Min | The number of completed ingestion batches. | Database, Table, Batching Type |
| Ingestion | BlobsDropped | Count | Sum, Max, Min | The number of blobs permanently dropped by a component, with each failure reason recorded in the IngestionResult metric. |
Database, Table, ComponentType, ComponentName |
| Ingestion | BlobsProcessed | Count | Sum, Max, Min | The number of blobs processed by a component. | Database, Table, ComponentType, ComponentName |
| Ingestion | BlobsReceived | Count | Sum, Max, Min | The number of blobs received from an input stream by a component. | Database, ComponentType, ComponentName |
| Export | ContinuousExportRecordsCount | Count | Sum | The number of exported records in all continuous export jobs. | Database, ContinuousExportName |
| Export | ContinuousExportMaxLateness | Count | Max | The lateness (minutes) reported by the continuous export jobs in the KQL Database. | |
| Export | ContinousExportPendingCount | Count | Max | The number of pending continuous export jobs that are ready to run but are waiting in a queue, possibly due to insufficient capacity. | |
| Export | ContinuousExportResult | The Failure/Success result of each continuous export run. | ContinuousExportName | The result of each continuous export run, indicating either failure or success. | ContinuousExportName |
| Ingestion | DiscoveryLatencyInSeconds | Seconds | Avg | The time from when data is enqueued until it's discovered by data connections. This time isn't included in the Stage latency or Ingestion latency metrics. Discovery latency might increase in the following situations: |
ComponentType, ComponentName |
| Ingestion | EventsDropped | Count | Sum, Max, Min | The number of events dropped by data connections. | ComponentType, ComponentName |
| Ingestion | EventsProcessed | Count | Sum, Max, Min | The number of events processed by data connections. | ComponentType, ComponentName |
| Ingestion | EventsReceived | Count | Sum, Max, Min | The number of events received by data connections from an input stream. | ComponentType, ComponentName |
| Ingestion | IngestionLatencyInSeconds | Seconds | Avg, Max, Min | The time taken from when data is received in the cluster until it's ready for query. The time depends on the ingestion type, such as Streaming Ingestion or Queued Ingestion. | IngestionKind |
| Ingestion | IngestionResult | Count | Sum | The total number of sources that were either successfully ingested or failed to be ingested. For more information, see Dimension descriptions | Database, Table, IngestionResultDetails, FailureKind, ViaUpdatePolicy |
| Ingestion | IngestionVolumeInBytes | Count | Max, Sum | The total size of data ingested to the KQL database (Bytes) before compression. | Database, Table |
| Materialized View | MaterializedViewAgeSeconds | Seconds | Avg | The age of the view (minutes) is defined by the current time minus the last ingestion time processed by the view. A lower value indicates a healthier view. | Database, MaterializedViewName |
| Materialized View | MaterializedViewHealth | 1, 0 | Avg | A value of 1 indicates the view is considered healthy; otherwise, the value is 0. | Database, MaterializedViewName |
| Materialized View | MaterializedViewResult | 1 | Avg | The metric value is always 1. Result indicates the result of the last materialization cycle. For possible values, see MaterializedViewResult. |
Database, MaterializedViewName, Result |
| Ingestion | QueueLength | Count | Avg | The number of pending messages in a component's input queue. The batching component processes one message per blob, while the ingestion component handles one message per batch. A batch consists of a single ingest command that includes one or more blobs. | ComponentType |
| Ingestion | QueueOldestMessage | Seconds | Avg | The time (seconds) from when the oldest message in a component's input queue was inserted. | ComponentType |
| Ingestion | ReceivedDataSizeBytes | Bytes | Avg, Sum | The size of the data received by data connections from an input stream. | ComponentType, ComponentName |
| StreamingIngestion | StreamingIngestDataRate | Bytes | Count, Avg, Max, Min, Sum | The total volume of data ingested by streaming ingestion. | Database, Table |
| StreamingIngestion | StreamingIngestDuration | Milliseconds | Avg, Max, Min | The total duration of all streaming ingestion requests. | None |
Dimension descriptions
The following list describes the dimensions reported in the IngestionResult metric:
IngestionResultDetails: Success for successful ingestion or the failure category for failures. For a complete list of possible failure categories, see Ingestion error codes.FailureKind: Whether the failure is permanent or transient. The value isNonefor a successful ingestion.ViaUpdatePolicy: True, if the ingestion was triggered by an Update Policy.
Note
- Event Hubs and IoT Hub ingestion events are pre-aggregated into one blob and then treated as a single ingestion source. They appear as a single ingestion result after pre-aggregation.
- Transient failures are automatically retried a limited number of times. Each transient failure is reported as a transient ingestion result, which means a single ingestion may generate multiple ingestion results.
Sample queries
You can find sample queries in the fabric-samples GitHub repository.