Monitor Cluster Metrics and Optimize Links for Cluster Linking on Confluent Platform

Confluent Platform exposes several metrics through Java Management Extensions (JMX) that are useful for monitoring Cluster Linking. Some of these are derived from extending existing interfaces, and others are new for Cluster Linking.

To monitor Cluster Linking, set the JMX_PORT environment variable before starting the cluster, then collect the reported metrics using your usual monitoring tools. JMXTrans, Graphite, and Grafana are a popular combination for collecting and reporting JMX metrics from Kafka. Datadog is another popular monitoring solution.

Quick Start for Administrators

• For basic health monitoring, focus on link-count (state=active) and mirror-topic-count metrics
• For performance monitoring, track MaxLag, BytesPerSec, and throttle time metrics
• For troubleshooting, monitor link-task-count (state=in-error) and connection failure metrics
• See the complete reference section for all available metrics

Also, you can set quotas to limit resources and bandwidth used by Cluster Linking to help optimize the system.

Monitoring brokers

Keep in mind that on KRaft these metrics apply only to brokers (not to the KRaft controller quorum). The broker that is the link coordinator will properly reflect the state of the link.

For example, the JMX MBean for Cluster Linking status is kafka.server:type=cluster-link-metrics,mode={mode},state={state},link-name={linkName}. The state can be active, paused, or unavailable. This MBean represents the total number of links in that specific state.

For the state=active MBean:

• Link coordinator broker: Returns 1 if the cluster link is active and managed by this broker
• Non-coordinator brokers: Return 0

Example: kafka.server:type=cluster-link-metrics,mode=source,state=active,link-name=my-link returns 1 on the link coordinator, 0 on other brokers.

In previous Confluent Platform releases that supported ZooKeeper (pre-8.0), the ZooKeeper “controller” broker acted as the link coordinator, so only that broker held the state of cluster links. The JMX MBean for Cluster Linking status against the “controller” would return 1 if the cluster link was active, but would return 0 on any other broker. ZooKeeper is no longer available on Confluent Platform 8.0 and later. To learn more, see KRaft and ZooKeeper in “What’s supported”.

Cluster link fetcher metrics

kafka.server.link:type=ClusterLinkFetcherManager,name=MaxLag,clientId=ClusterLink,link-name={linkName}

Maximum lag in messages between the replicas on the destination cluster and the leader replica on the source cluster. Provides an indication of overall cluster link health and a relative measure of whether the destination cluster is lagging behind.

kafka.server.link:type=ClusterLinkFetcherManager,name=FetchTimeMax,clientId=ClusterLink,link-name={linkName},state={state}

This metric shows the longest time gap between successful data fetches for a partition. The clock starts when a partition is ready to be used and stops when a successful fetch occurs (even if no new data comes in). If the time since the last fetch is longer than the delay from the previous fetch, that longer period is used. Partitions with unavailable, failed, or degraded links are not counted. Normally, the fetch interval is low. If you notice a consistently high interval, it could signal an underlying issue.

Available tags: - state: The fetch state (fetch for active fetching, other values for different states)

kafka.server:type=FetcherLagMetrics,name=ConsumerLag,clientId=ClusterLinkFetcherThread-{destBrokerId}-{linkName}-{sourceBrokerId},topic={topic},partition={partition},link-name={linkName}

Drill-down of MaxLag, which provides lag in number of messages per replica on the destination cluster from the leader replica on the source cluster. Useful to debug any specific lagging topic partitions/replicas on the destination cluster.

kafka.server.link:type=ClusterLinkFetcherManager,name=DeadThreadCount,clientId=ClusterLink,link-name={linkName}

Number of dead threads of type ClusterLinkThread, where a dead thread is indicative of an issue with fetching from the source cluster for a partition.

kafka.server.link:type=ClusterLinkFetcherManager,name=DegradedPartitionCount,clientId=ClusterLink,link-name={linkName}

Partitions that have been moved to degraded state, typically due to a non-critical error after retry timeout. These include user errors, such as authorization failure and includes link availability failures, indicated by the reason tag. A sustained value greater than (>) 0 beyond multiple metadata intervals and availability timeout may indicate an issue, depicted by the reason tag internal.

kafka.server:type=FetcherStats,name=BytesPerSec,clientId=ClusterLinkFetcherThread-{destBrokerId}-{linkName}-{sourceBrokerId},brokerHost={host},brokerPort={port},link-name={linkName}

Rate at which data is fetched from the source cluster. Indicates amount of throughput in bytes per second on the cluster link.

kafka.server:type=FetcherStats,name=RequestsPerSec,clientId=ClusterLinkFetcherThread-{destBrokerId}-{linkName}-{sourceBrokerId}, brokerHost={host},brokerPort={port},link-name={linkName}

Rate at which requests are issued to the source cluster over the cluster link.

Throttle time metrics for cluster link fetchers

kafka.server:type=cluster-link,link-name={linkName}

Throttle time metrics may indicate that a cluster link connection is being throttled, which is useful for understanding why lag may be increasing over a cluster link.

Name	Description
fetch-throttle-time-avg, fetch-throttle-time-max	Gives throttle times for the Cluster Linking fetchers. May indicate increases in lag on the cluster link due to throttling/quotas being enforced.

Cluster link network client metrics

kafka.server:type=cluster-link-metadata-metrics,link-name={linkName}

Metrics pertaining to the metadata refresh client in the cluster link.

kafka.server:type=cluster-link-fetcher-metrics,link-name={linkName},broker-id={id},fetcher-id={id}

Metrics pertaining to data fetcher requests in the cluster link.

Metrics seen through cluster-link-fetcher-metrics are shown in the following table.

Name	Description
connection-count	Number of connections for a cluster link
connection-creation-rate, connection-creation-total	Rate per second and total for connections created for the cluster link. If the rate is high, the source cluster may be overloaded with connections.
connection-close-rate, connection-close-total	Rate per second and total for connections closed for the cluster link. Can be compared to connection creation to understand the balance of creating/closing connections.
incoming-byte-rate, incoming-byte-total	Number of bytes per second and total received from the source cluster.
outgoing-byte-rate, outgoing-byte-total	Number of bytes per second and total sent to the source cluster.
network-io-rate, network-io-total	Indicates rate and total network input/output (IO).
request-rate, request-total	Rate per second and total requests issued to the source cluster over the cluster link.
response-rate, response-total	Rate per second and total number of responses received from the source cluster over the cluster link.
request-size-avg, request-size-max	Average and maximum size in bytes of requests issued to the source cluster over the cluster link.
io-ratio, io-wait-time-ns-avg, io-waittime-total, iotime-total	Statistics of the destination cluster’s network IO for the requests over the cluster link.
successful-authentication-rate, successful-authentication-total	Rate per second and total number of cluster link clients authenticating to the source cluster over the cluster link.
successful-reauthentication-rate, successful-reauthentication-total	Rate per second and total re-authentication to the source cluster over the cluster link.
failed-reauthentication-rate, failed-reauthentication-total	Rate per second and total failed re-authentications to the source cluster over the cluster link. If failures are present, it could indicate misconfigured or stale cluster link credentials.
reauthentication-latency-avg	Average re-authentication latency. Helps to assess whether clients are taking too long to authenticate to the clusters.

Request metrics for cluster link APIs

kafka.network:type=RequestMetrics,name={ LocalTimeMs| RemoteTimeMs | RequestBytes | RequestQueueTimeMs| ResponseQueueTimeMs | ResponseSendIoTimeMs | ResponseSendTimeMs | ThrottleTimeMs| TotalTimeMs },request={CreateClusterLinks| DeleteClusterLinks| ListClusterLinks}

Depending on the request name, provides statistics on requests on the cluster link, including requests and response times, time requests wait in the queue, size of requests (bytes), and so forth.

kafka.server:type=DelayedOperationPurgatory,name=PurgatorySize,delayedOperation=ClusterLink

Provides metrics on delayed operations on the cluster link.

Task status metrics

kafka.server:type=cluster-link-metrics,name=link-task-count,link-name={linkName},task-name={taskName},state={state},reason={reason},mode={mode},connection-mode={connection_mode}

Monitor the state of link level tasks. For example, monitor if consumer offset syncing is working. If the task is in error, a reason code is provided. You can set up alerts to trigger if errors occur.

Available tags:

• task-name: The specific task being monitored. Possible values:
  • consumer-offset-sync: Consumer offset synchronization task
  • acl-sync: ACL synchronization task
  • auto-create-mirror: Automatic mirror topic creation task
  • topic-configs-sync: Topic configuration synchronization task
  • clear-mirror-start-offsets: Clear mirror start offsets task
  • pause-mirror-topics: Pause mirror topics task
  • check-availability: Availability check task
  • state-aggregator: State aggregation task
  • retry-task: Retry task for failed operations
  • periodic-partition-scheduler: Periodic partition scheduler task
  • degraded-partition-monitor: Degraded partition monitor task
• state: The current state of the task. Possible values:
  • active: Task is currently running
  • in-error: Task has encountered an error
• reason: Error code when the task is in error state. Common values include:
  • no-error: No error (when state is active)
  • authentication: Authentication errors with link credentials [Link cannot authenticate to remote cluster]
• broker-authentication: Authentication errors with broker credentials [Authentication issues between link and destination broker]
• authorization: Authorization errors with link credentials [Link lacks permissions on remote cluster]
• broker-authorization: Authorization errors with broker credentials [Authorization issues between link and destination broker]
• misconfiguration: Configuration errors [Invalid or missing link configuration]
• internal: Internal/unexpected errors [Unexpected system errors]
• suppressed-errors: Errors that are being suppressed [Errors being handled gracefully]
• consumer-group-in-use: Consumer group is active on destination [Cannot sync offsets while consumers are active]
• remote-link-not-found: Remote link not found (bidirectional links) [Remote side of bidirectional link missing]
• security-disabled: Remote cluster has no authorizer configured [Remote cluster lacks security configuration]
• acl-limit-exceeded: ACL limit reached on destination cluster [ACL quota exceeded on destination]
• invalid-request: Invalid request error [Malformed or invalid API request]
• topic-exists: Topic already exists on destination [Cannot create mirror topic - already exists]
• policy-violation: Policy violation error [Mirror topic creation violates policies]
• invalid-topic: Invalid topic error [Topic name or configuration invalid]
• unknown-topic-or-partition: Topic or partition not found [Source topic/partition doesn’t exist]

kafka.server:type=cluster-link-metrics,name=mirror-transition-in-error,link-name={linkName},state={state},reason={reason},mode={mode},connection-mode={connection_mode}

Monitor mirror topic state transition errors. For example, if a mirror topic encounters errors during the promotion process; that is, while its state is pending_stopped and it is being transitioned to stopped.

Available tags:

• state: The mirror topic transition state. Common values include:
  • pending_stopped: Mirror topic is being stopped
  • pending_mirror: Topic is being converted to mirror
  • pending_synchronization: Mirror topic is being reversed/swapped
  • pending_restore: Mirror topic is being prepared for restore
  • pending_setup_for_restore: Mirror topic is being prepared for restore setup
  • failed: Mirror topic repair operations
• reason: Error code for the transition failure. Uses the same error codes as task metrics (see above).

Metrics on the destination cluster

The following broker metrics, specific to Cluster Linking, are available on the brokers in the destination cluster.

kafka.server:type=cluster-link-metrics,name=mirror-partition-count,link-name={linkName}

Number of actively mirrored partitions for which the broker is leader

kafka.server.link:type=ClusterLinkFetcherManager,name=FailedPartitionsCount,clientId=ClusterLink,link-name={linkName}

Number of failed mirror partitions for which the broker is leader. A partition may be in a failed state if the source partition’s epoch has gone backwards; for example, if the topic was recreated.

kafka.server:type=ReplicaManager,name=UnderMinIsrMirrorPartitionCount

Number of mirrored partitions that are under min ISR.

kafka.server:type=ReplicaManager,name=UnderReplicatedMirrorPartitions

Number of mirrored partitions that are under replicated.

kafka.server:type=BrokerTopicMetrics,name=MirrorBytesInPerSec,topic={topic}

Rate of bytes per second written to mirror topics through Cluster Linking. This metric is available both per-topic (with topic tag) and as an aggregate across all mirror topics (without topic tag). Available since Confluent Platform 7.6.1.

This metric specifically tracks data written to the destination cluster’s mirror topics, which complements the existing FetcherStats metrics that track overall cluster link throughput. Use this metric for monitoring mirror topic write performance and comparing it with source topic throughput.

Available tags:

• topic: The mirror topic name (present for per-topic metrics, absent for aggregate metrics)

kafka.server:type=cluster-link-metrics,name=(linked-leader-epoch-change-rate, linked-leader-epoch-change-total),link-name={linkName}

Rate per second and total number of times leader election was triggered on this broker due to source leader changes.

• Frequent triggers for leader election might indicate issues on the source cluster. This can be a useful metric during on-premises to Confluent Cloud migrations to identify if there are issues on the source cluster.

kafka.server:type=cluster-link-metrics,name=(linked-topic-partition-addition-rate, linked-topic-partition-addition-total),link-name={linkName}

Rate per second and total number of times at which the partition count was updated due to source changes.

• A high volume or rate of partition count updates might indicate issues on the source cluster.

kafka.server:type=cluster-link-metrics,state=Mirror,link-name={link_name},name=mirror-topic-count

Total number of active (healthy) mirror topics.

Note

Note: Known issue: This metric will include topics that are in the SOURCE_UNAVAILABLE state. As a temporary workaround, use kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},state=active,link-name={linkName} and kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},state=unavailable,link-name={linkName} to alert on the source absence.

kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},state=active,link-name={linkName}

Total number of active cluster links connected to the cluster. You can filter or group by direction (source or destination) and connection mode (inbound or outbound).

kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},state=paused,link-name={linkName}

Total number of paused cluster links connected to the cluster. You can filter or group by direction (source or destination) and connection mode (inbound or outbound).

kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},state=unavailable,link-name={linkName},reason={reason}

Total number of unavailable cluster links connected to the cluster. You can filter or group by direction (source or destination) and connection mode (inbound or outbound). The reason tag provides specific details about why the link is unavailable.

kafka.server:type=cluster-link-metrics,state=PausedMirror,link-name={link_name},name=mirror-topic-count

Total number of mirror topics for which mirroring from the source has been paused.

kafka.server:type=cluster-link-metrics,state=PendingStoppedMirror,link-name={link_name},name=mirror-topic-count

Total number of mirror topics whose mirroring has been temporarily stopped.

kafka.server:type=cluster-link-metrics,state=FailedMirror,link-name={link_name},name=mirror-topic-count

Total number of failed mirror topics; that is, mirroring cannot proceed.

kafka.server:type=cluster-link-metrics,state=active,link-name={link_name},mode={mode},connection-mode={connection_mode},name=link-count

Total number of links in an active state.

Available tags:

• mode: The direction of the cluster link. Possible values:
  • destination: Links where this cluster acts as the destination
  • source: Links where this cluster acts as the source
  • bidirectional: Links that can operate in both directions
• connection-mode: The connection mode for the cluster link. Possible values:
  • inbound: Connections coming into this cluster
  • outbound: Connections going out from this cluster

kafka.server:type=cluster-link-metrics,state=paused,link-name={link_name},mode={mode},connection-mode={connection_mode},name=link-count

Total number of links in a paused state.

kafka.server:type=cluster-link-metrics,state=unavailable,link-name={link_name},mode={mode},connection-mode={connection_mode},reason={reason},name=link-count

Total number of links in an unavailable state, where the source may not be available the link is having trouble fetching.

Additional tags for unavailable links:

• reason: The specific reason why the link is unavailable. Common values include authentication failures, authorization issues, network connectivity problems, and internal errors.

kafka.server:type=cluster-link-metrics,link-name={link_name},name=broker-failed-link-count

Total number of links in a failed state. This metric is reported on a per broker state, as some or all brokers could have a link in a failed state and unable to connect to the source.

kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},remote-link-connection-mode={remote_connection_mode},link-name={link_name},name=controller-reverse-connection-count

Number of persistent reverse connections for this link between the local broker and the remote link coordinator.

Available tags:

• remote-link-connection-mode: The connection mode on the remote side. Possible values:
  • inbound: Remote cluster accepts connections from this cluster
  • outbound: Remote cluster initiates connections to this cluster

kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},link-name={link_name},name=reverse-connection-count

Total number of reverse connections for this link between local and remote brokers.

kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},link-name={link_name},name=reverse-connection-created-total,reverse-connection-created-rate

Total count and rate per second of reverse connections created for this cluster link.

kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},link-name={link_name},name=reverse-connection-closed-total,reverse-connection-closed-rate

Total count and rate per second of reverse connections closed for this cluster link.

kafka.server:type=cluster-link-metrics,mode={mode},connection-mode={connection_mode},link-name={link_name},name=reverse-connection-failed-total,reverse-connection-failed-rate

Total count and rate per second of reverse connections that failed for this cluster link. Only available for outbound connections.

kafka.server:type=cluster-link-metrics,link-name={link_name},name=link-fetcher-count

Number of link fetchers currently active for this cluster link.

kafka.server:type=cluster-link-metrics,pool={pool},link-name={link_name},name=link-target-fetcher-count

Number of cluster link target fetchers per source broker.

Available tags:

• pool: The fetcher pool type. Possible values include different fetcher pool configurations.

kafka.server:type=cluster-link-metrics,link-name={link_name},name=link-fetcher-throttled-partition-count

Number of throttled partitions for this cluster link.

kafka.server:type=cluster-link-metrics,link-name={link_name},name=link-fetcher-unassigned-partition-count

Number of unassigned partitions for this cluster link.

kafka.server:type=cluster-link-metrics,reason={reason},link-name={link_name},name=link-fetcher-degraded-partition-count

Number of partitions in degraded state due to specific reasons.

Available tags:

• reason: The reason for partition degradation. Specific degradation reasons include various operational issues.

kafka.server:type=cluster-link-metrics,link-name={link_name},name=remote-admin-queue-size

Backlog queue size for the remote admin client for this cluster link.

Metrics on the source cluster

The source cluster (with Kafka or Confluent Server previous to 6.1.0) is unaware of cluster links, but can monitor cluster link related usage by allocating link-specific credentials. Quota metrics for each credential is available in both Kafka and Confluent Platform when user quotas are enabled.

kafka.server:type=Fetch,user={linkPrincipal},client-id={optionalLinkClientId}

• byte-rate
• throttle-time

kafka.server:type=Request,user={linkPrincipal},client-id={optionalLinkClientId}

• request-time
• throttle-time

If the source cluster is Confluent Server with a version of 6.1.0 or higher, then you can also monitor cluster links on the source cluster with the following metrics.

kafka.server:type=cluster-link-source-metrics,request={request},link-id={linkUUID}

• request-byte-total
• request-total
• response-byte-total
• response-time-ns-avg

Setting quotas on resources and bandwidth usage

You can set various types of quotas to place caps on resources and bandwidth used for Cluster Linking on the source and destination clusters.

Destination cluster quotas

You can limit total usage for cluster links on each broker in the destination cluster by setting the broker config confluent.cluster.link.io.max.bytes.per.second.

Source cluster quotas

The source cluster is unaware of cluster links, but can limit usage for cluster links by allocating link-specific credentials and assigning quotas for the link principal.

• Fetch byte-rate (replication throughput) for each cluster link principal
• Request rate quota (CPU/thread usage) for each cluster link principal. This also includes CPU usage for metadata and configuration sync as well as ACL and consumer offset migration.

Monitoring Use Cases and Alert Recommendations

Here are common monitoring scenarios and recommended metrics for cluster link operations:

Health Monitoring

• Link Status: kafka.server:type=cluster-link-metrics,name=link-count,state=active - Should be 1 for healthy links
• Mirror Topic Health: kafka.server:type=cluster-link-metrics,name=mirror-topic-count,state=Mirror - Track active mirror topics
• Failed Links: kafka.server:type=cluster-link-metrics,name=broker-failed-link-count - Alert when > 0

Performance Monitoring

• Replication Lag: kafka.server.link:type=ClusterLinkFetcherManager,name=MaxLag - Monitor cross-cluster lag
• Throughput: kafka.server:type=FetcherStats,name=BytesPerSec - Track data transfer rates
• Mirror Topic Throughput: kafka.server:type=BrokerTopicMetrics,name=MirrorBytesInPerSec - Track mirror topic write rates (CP 7.6.1+)
• Throttling: kafka.server:type=cluster-link,name=fetch-throttle-time-avg - Monitor throttling impact

Error Monitoring

• Task Failures: kafka.server:type=cluster-link-metrics,name=link-task-count,state=in-error - Alert on task errors
• Connection Issues: kafka.server:type=cluster-link-metrics,name=reverse-connection-failed-total - Monitor connection failures
• Authentication Problems: Use reason=authentication or reason=authorization tags to identify auth issues

Capacity Planning

• Connection Count: kafka.server:type=cluster-link-metrics,name=reverse-connection-count - Track connection usage
• Queue Sizes: kafka.server:type=cluster-link-metrics,name=remote-admin-queue-size - Monitor queue backlogs
• Thread Usage: kafka.server:type=cluster-link-metrics,name=background-thread-usage - Track resource utilization

Complete Metrics and Tags Reference

This section provides a comprehensive reference for all Cluster Linking metrics and their available tags.

Common Tags Used Across Multiple Metrics

Core Tags:

• link-name={linkName}: The name of the cluster link
• mode={mode}: Link direction (destination, source, bidirectional)
• connection-mode={connectionMode}: Connection mode (inbound, outbound)
• state={state}: Metric-specific state values
• reason={reason}: Error codes or specific reasons

Specialized Tags:

• remote-link-connection-mode={remoteConnectionMode}: Remote side connection mode
• task-name={taskName}: Specific task being monitored
• pool={pool}: Fetcher pool type
• thread={threadId}: Background thread identifier

Link Status and Count Metrics

Metric Name	Tags	Description
link-count	link-name, state, mode, connection-mode, reason (unavailable only)	Number of links in specified state
broker-failed-link-count	link-name	Links in failed state per broker
remote-link-state-count	link-name, remote-link-state, mode, connection-mode	Count by remote link state
unavailable-link-count	link-name, reason, mode, connection-mode	Unavailable links with reason
intranet-connectivity-link-count	link-name, mode, connection-mode	Links using intranet connectivity
prefixed-destination-link-count	None	Links with prefix enabled

Mirror Topic and Partition Metrics

Metric Name	Tags	Description
mirror-partition-count	link-name	Active mirror partitions
mirror-topic-count	link-name, state	Mirror topics by state
failed-mirror-topic-count	link-name, reason, mode, connection-mode	Failed mirror topics with reason
MirrorBytesInPerSec	topic (optional)	Bytes per second written to mirror topics (CP 7.6.1+)

Connection and Network Metrics

Metric Name	Tags	Description
controller-reverse-connection-count	link-name, mode, connection-mode, remote-link-connection-mode	Persistent controller connections
reverse-connection-count	link-name, mode, connection-mode	Total reverse connections
reverse-connection-created-total|rate	link-name, mode, connection-mode	Reverse connections created
reverse-connection-closed-total|rate	link-name, mode, connection-mode	Reverse connections closed
reverse-connection-failed-total|rate	link-name, mode, connection-mode	Failed reverse connections

Fetcher and Performance Metrics

Metric Name	Tags	Description
link-fetcher-count	link-name	Active link fetchers
link-target-fetcher-count	link-name, pool	Target fetchers per pool
link-fetcher-throttled-partition-count	link-name	Throttled partitions
link-fetcher-unassigned-partition-count	link-name	Unassigned partitions
link-fetcher-degraded-partition-count	link-name, reason	Degraded partitions with reason
fetch-throttle-time-avg|max	link-name	Fetch throttle time statistics

Task Status Metrics

Metric Name	Tags	Description
link-task-count	link-name, task-name, state, reason, mode, connection-mode	Task status monitoring
mirror-transition-in-error	link-name, state, reason, mode, connection-mode	Mirror transition errors

Administrative and Queue Metrics

Metric Name	Tags	Description
remote-admin-queue-size	mode, connection-mode	Remote admin queue backlog
remote-admin-queue-time-ms-avg|max	link-name	Admin queue wait time
remote-admin-request-time-ms-avg|max	link-name	Admin request processing time
local-admin-queue-size	thread	Local admin queue per thread
local-admin-queue-time-ms-avg|max	link-name	Local admin queue wait time
local-admin-request-time-ms-avg|max	link-name	Local admin request processing time

Data Sync and Topic Management Metrics

Metric Name	Tags	Description
linked-topic-partition-addition-total|rate	link-name	Topic partition additions
linked-leader-epoch-change-total|rate	link-name	Leader epoch changes
acls-added-total|rate	link-name	ACLs successfully added
acls-add-failed-total|rate	link-name	Failed ACL additions
acls-deleted-total|rate	link-name	ACLs successfully deleted
acls-delete-failed-total|rate	link-name	Failed ACL deletions
describe-acls-from-source-failed-total|rate	link-name	Failed ACL descriptions
consumer-offset-committed-total|rate	link-name	Consumer offsets synced
consumer-offset-commit-failed-total|rate	link-name	Failed offset commits
topic-config-update-total|rate	link-name	Topic config updates
topic-config-update-failed-total|rate	link-name	Failed config updates

Auto-Mirror and Automation Metrics

Metric Name	Tags	Description
auto-mirror-created-total|rate	link-name	Auto-created mirror topics
auto-mirror-create-failed-total|rate	link-name	Failed auto-mirror creation
auto-mirror-list-topics-from-source-failed-total|rate	link-name	Failed source topic listing
auto-mirror-list-topics-from-destination-failed-total|rate	link-name	Failed destination topic listing
auto-mirror-list-mirrors-from-source-failed-total|rate	link-name	Failed mirror listing from source
list-consumer-group-offsets-from-source-failed-total|rate	link-name	Failed offset listing from source
list-consumer-group-offsets-from-destination-failed-total|rate	link-name	Failed offset listing from destination
list-consumer-groups-from-source-failed-total|rate	link-name	Failed consumer group listing
prefixed-auto-mirror-created-total|rate	link-name	Prefixed auto-created mirrors
prefixed-auto-mirror-create-failed-total|rate	link-name	Failed prefixed auto-creation
prefixed-auto-mirror-topic-filtered-count	link-name	Topics filtered out from mirroring

Throttle and Performance Timing Metrics

Metric Name	Tags	Description
destination-lag-link-fetcher-throttle-total|rate	link-name	Throttle due to destination lag
link-fetcher-produce-throttle-total|rate	link-name	Produce quota throttling
link-fetcher-request-throttle-total|rate	link-name	Request quota throttling
link-fetcher-fetch-time-avg|max	link-name, state	Fetch timing (state=fetch or wait)
time-to-stop-mirror-topic-failover-ms-avg|max	link-name	Time to failover mirror topic
time-to-stop-mirror-topic-promote-ms-avg|max	link-name	Time to promote mirror topic
link-unavailable-total|rate	link-name, mode, connection-mode	Link unavailability incidents

Background Thread and Resource Metrics

Metric Name	Tags	Description
background-thread-usage	thread	Background thread active usage
background-thread-tenants	thread	Tenants per background thread
background-thread-link-coordinators	thread	Link coordinators per thread
background-thread-max-task-wait-time-ms	thread	Max task wait time per thread
fetch-size-allocated-total	pool	Total fetch size allocated
tenant-fetch-size-allocated-min	pool	Min tenant fetch allocation
tenant-fetch-size-allocated-max	pool	Max tenant fetch allocation

Multi-Tenant Specific Metrics

Metric Name	Tags	Description
metadata-topic-owned-partitions-count	None	Metadata topic partitions owned
link-count-with-owned-link-coordinator	None	Links with owned coordinator
max-link-count-per-metadata-topic-partition	None	Max links per metadata partition

Task Names for task-name Tag

• consumer-offset-sync: Consumer offset synchronization
• acl-sync: ACL synchronization
• auto-create-mirror: Automatic mirror creation
• topic-configs-sync: Topic configuration sync
• clear-mirror-start-offsets: Clear mirror start offsets
• pause-mirror-topics: Pause mirror topics
• check-availability: Availability checks
• state-aggregator: State aggregation
• retry-task: Retry failed operations
• periodic-partition-scheduler: Partition scheduler
• degraded-partition-monitor: Degraded partition monitor

Common Error Codes for reason Tag

Authentication/Authorization:

• authentication, broker-authentication
• authorization, broker-authorization
• security-disabled

Configuration Issues:

• misconfiguration
• policy-violation
• invalid-request

Resource Issues:

• topic-exists
• invalid-topic
• unknown-topic-or-partition
• acl-limit-exceeded
• consumer-group-in-use

Operational Issues:

• internal
• suppressed-errors
• remote-link-not-found
• remote-mirror-not-found
• link-not-found

State Transition Issues:

• unexpected-mirror-state
• failed-to-update-metadata
• failover-timeout

Related content

• See Monitoring Kafka with JMX in Confluent Platform for a full list of metrics for Confluent Platform.
• See the Apache Kafka® documentation on metrics reporting and monitoring through JMX endpoints.