Supported metrics for Microsoft.Kusto/clusters

The following table lists the metrics available for the Microsoft.Kusto/clusters resource type.

Table headings

Metric - The metric display name as it appears in the Azure portal.
Name in Rest API - Metric name as referred to in the REST API.
Unit - Unit of measure.
Aggregation - The default aggregation type. Valid values: Average, Minimum, Maximum, Total, Count.
Dimensions - Dimensions available for the metric.
Time Grains - Intervals at which the metric is sampled. For example, PT1M indicates that the metric is sampled every minute, PT30M every 30 minutes, PT1H every hour, and so on.
DS Export- Whether the metric is exportable to Azure Monitor Logs via Diagnostic Settings.

For information on exporting metrics, see Create diagnostic settings in Azure Monitor.

For information on metric retention, see Azure Monitor Metrics overview.

For a list of supported logs, see Supported log categories - Microsoft.Kusto/clusters

Category Metric Name in REST API Unit Aggregation Dimensions Time Grains DS Export
Ingestion health and performance Batch Blob Count

Number of data sources in an aggregated batch for ingestion.
BatchBlobCount Count Average, Maximum, Minimum Database PT1M Yes
Ingestion health and performance Batch Duration

The duration of the aggregation phase in the ingestion flow.
BatchDuration Seconds Average, Maximum, Minimum Database PT1M Yes
Ingestion health and performance Batches Processed

Number of batches aggregated for ingestion. Batching Type: whether the batch reached batching time, data size or number of files limit set by batching policy
BatchesProcessed Count Total (Sum), Average, Maximum, Minimum Database, SealReason PT1M Yes
Ingestion health and performance Batch Size

Uncompressed expected data size in an aggregated batch for ingestion.
BatchSize Bytes Average, Maximum, Minimum Database PT1M Yes
Ingestion health and performance Blobs Dropped

Number of blobs permanently rejected by a component.
BlobsDropped Count Total (Sum), Average, Minimum, Maximum Database, ComponentType, ComponentName PT1M Yes
Ingestion health and performance Blobs Processed

Number of blobs processed by a component.
BlobsProcessed Count Total (Sum), Average, Minimum, Maximum Database, ComponentType, ComponentName PT1M Yes
Ingestion health and performance Blobs Received

Number of blobs received from input stream by a component.
BlobsReceived Count Total (Sum), Average, Minimum, Maximum Database, ComponentType, ComponentName PT1M Yes
Cluster health Cache utilization (deprecated)

Utilization level in the cluster scope. The metric is deprecated and presented for backward compatibility only, you should use the 'Cache utilization factor' metric instead.
CacheUtilization Percent Average, Maximum, Minimum <none> PT1M Yes
Cluster health Cache utilization factor

Percentage of utilized disk space dedicated for hot cache in the cluster. 100% means that the disk space assigned to hot data is optimally utilized. No action is needed in terms of the cache size. More than 100% means that the cluster's disk space is not large enough to accommodate the hot data, as defined by your caching policies. To ensure that sufficient space is available for all the hot data, the amount of hot data needs to be reduced or the cluster needs to be scaled out. Enabling auto scale is recommended.
CacheUtilizationFactor Percent Average, Maximum, Minimum <none> PT1M Yes
Export health and performance Continuous Export Max Lateness

The lateness (in minutes) reported by the continuous export jobs in the cluster
ContinuousExportMaxLatenessMinutes Count Maximum <none> PT1M Yes
Export health and performance Continuous export - num of exported records

Number of records exported, fired for every storage artifact written during the export operation
ContinuousExportNumOfRecordsExported Count Total (Sum) ContinuousExportName, Database PT1M Yes
Export health and performance Continuous Export Pending Count

The number of pending continuous export jobs ready for execution
ContinuousExportPendingCount Count Maximum <none> PT1M Yes
Export health and performance Continuous Export Result

Indicates whether Continuous Export succeeded or failed
ContinuousExportResult Count Count ContinuousExportName, Result, Database PT1M Yes
Cluster health CPU

CPU utilization level
CPU Percent Average, Maximum, Minimum <none> PT1M Yes
Ingestion health and performance Discovery Latency

Reported by data connections (if exist). Time in seconds from when a message is enqueued or event is created until it is discovered by data connection. This time is not included in the Azure Data Explorer total ingestion duration.
DiscoveryLatency Seconds Average ComponentType, ComponentName PT1M Yes
Ingestion health and performance Events Dropped

Number of events dropped permanently by data connection. An Ingestion result metric with a failure reason will be sent.
EventsDropped Count Total (Sum), Average, Minimum, Maximum ComponentType, ComponentName PT1M Yes
Ingestion health and performance Events Processed

Number of events processed by the cluster
EventsProcessed Count Total (Sum), Average, Minimum, Maximum ComponentType, ComponentName PT1M Yes
Ingestion health and performance Events Received

Number of events received by data connection.
EventsReceived Count Total (Sum), Average, Minimum, Maximum ComponentType, ComponentName PT1M Yes
Export health and performance Export Utilization

Export utilization
ExportUtilization Percent Maximum <none> PT1M Yes
Cluster health FollowerLatency

The follower databases synchronize changes in the leader databases. Because of the synchronization, there's a data lag of a few seconds to a few minutes in data availability.This metric measures the length of the time lag. The time lag depends on the overall size of the leader database metadata.This is a cluster level metrics: the followers catch metadata of all databases that are followed. This metric represents the latency of the process.
FollowerLatency MilliSeconds Average, Maximum, Minimum State, RoleInstance PT1M Yes
Ingestion health and performance Ingestion Latency

Latency of data ingested, from the time the data was received in the cluster until it's ready for query. The ingestion latency period depends on the ingestion scenario.
IngestionLatencyInSeconds Seconds Average, Maximum, Minimum IngestionKind PT1M Yes
Ingestion health and performance Ingestion result

Total number of sources that either failed or succeeded to be ingested. Splitting the metric by status, you can get detailed information about the status of the ingestion operations.
IngestionResult Count Total (Sum) IngestionResultDetails, FailureKind PT1M Yes
Cluster health Ingestion utilization

Ratio of used ingestion slots in the cluster
IngestionUtilization Percent Average, Maximum, Minimum <none> PT1M Yes
Ingestion health and performance Ingestion Volume

Overall volume of ingested data to the cluster
IngestionVolumeInMB Bytes Total (Sum), Maximum Database PT1M Yes
Cluster health Instance Count

Total instance count
InstanceCount Count Average, Maximum, Minimum, Count, Total (Sum) <none> PT1M Yes
Cluster health Keep alive

Sanity check indicates the cluster responds to queries
KeepAlive Count Average <none> PT1M Yes
Materialized View health and performance Materialized View Age

The materialized view age in minutes
MaterializedViewAgeMinutes Count Average Database, MaterializedViewName PT1M Yes
Materialized View health and performance Materialized View Age

The materialized view age in seconds
MaterializedViewAgeSeconds Seconds Average, Minimum, Maximum Database, MaterializedViewName PT1M Yes
Materialized View health and performance Materialized View Data Loss

Indicates potential data loss in materialized view
MaterializedViewDataLoss Count Maximum Database, MaterializedViewName, Kind PT1M Yes
Materialized View health and performance Materialized View Extents Rebuild

Number of extents rebuild
MaterializedViewExtentsRebuild Count Average Database, MaterializedViewName PT1M Yes
Materialized View health and performance Materialized View Health

The health of the materialized view (1 for healthy, 0 for non-healthy)
MaterializedViewHealth Count Average Database, MaterializedViewName PT1M Yes
Materialized View health and performance Materialized View Records In Delta

The number of records in the non-materialized part of the view
MaterializedViewRecordsInDelta Count Average Database, MaterializedViewName PT1M Yes
Materialized View health and performance Materialized View Result

The result of the materialization process
MaterializedViewResult Count Average Database, MaterializedViewName, Result PT1M Yes
Partitioning Partitioning Percentage

Percentage of records partitioned versus total number of records.
PartitioningPercentage Percent Average, Maximum, Minimum Database, Table PT1M Yes
Partitioning Partitioning Percentage Hot

Percentage of records partitioned versus total number of records (in hot / cached extents only).
PartitioningPercentageHot Percent Average, Maximum, Minimum Database, Table PT1M Yes
Partitioning Processed Partitioned Records

Number of records partitioned in measured time window.
ProcessedPartitionedRecords Count Average, Maximum, Minimum Database, Table PT1M Yes
Query performance Query duration

Queries duration in seconds
QueryDuration MilliSeconds Average, Maximum, Minimum, Total (Sum) QueryStatus PT1M Yes
Query performance Query Result

Total number of queries.
QueryResult Count Count QueryStatus PT1M No
Ingestion health and performance Queue Length

Number of pending messages in a component's queue.
QueueLength Count Average ComponentType PT1M Yes
Ingestion health and performance Queue Oldest Message

Time in seconds from when the oldest message in queue was inserted.
QueueOldestMessage Count Average ComponentType PT1M Yes
Ingestion health and performance Received Data Size Bytes

Size of data received by data connection. This is the size of the data stream, or of raw data size if provided.
ReceivedDataSizeBytes Bytes Average, Total (Sum) ComponentType, ComponentName PT1M Yes
Ingestion health and performance Stage Latency

Cumulative time from when a message is discovered until it is received by the reporting component for processing (discovery time is set when message is enqueued for ingestion queue, or when discovered by data connection).
StageLatency Seconds Average Database, ComponentType PT1M Yes
Streaming Ingest Streaming Ingest Data Rate

Streaming ingest data rate
StreamingIngestDataRate Bytes Average, Minimum, Maximum <none> PT1M Yes
Streaming Ingest Streaming Ingest Duration

Streaming ingest duration in milliseconds
StreamingIngestDuration MilliSeconds Average, Minimum, Maximum <none> PT1M Yes
Streaming Ingest Streaming Ingest Result

Streaming ingest result
StreamingIngestResults Count Count Result PT1M Yes
Streaming Ingest Streaming Ingest Utilization

Streaming Ingest Utilization is the percentage of actual concurrent streaming ingestion requests performed, compared to the maximum number of concurrent streaming ingestion requests.
StreamingIngestUtilization Percent Average, Maximum, Minimum <none> PT1M Yes
Query performance Total number of concurrent queries

Total number of concurrent queries
TotalNumberOfConcurrentQueries Count Average, Maximum, Minimum <none> PT1M Yes
Cluster health Total number of extents

Total number of data extents
TotalNumberOfExtents Count Average, Maximum, Minimum <none> PT1M Yes
Cluster health Total number of throttled commands

Total number of throttled commands
TotalNumberOfThrottledCommands Count Average, Maximum, Minimum, Total (Sum) CommandType PT1M Yes
Query performance Total number of throttled queries

Total number of throttled queries
TotalNumberOfThrottledQueries Count Average, Maximum, Minimum, Total (Sum) <none> PT1M Yes
Query performance Weak consistency latency

The max latency between the previous metadata sync and the next one (in DB/node scope)
WeakConsistencyLatency Seconds Average, Maximum, Minimum Database, RoleInstance PT1M Yes

Next steps