Metrics Reference
This section provides reference descriptions for the public metrics exported from Redpanda’s /public_metrics endpoint.
|
In a live system, Redpanda metrics are exported only for features that are in use. For example, Redpanda does not export metrics for consumer groups if no groups are registered. To see the available public metrics in your system, query the bash |
Cluster metrics
redpanda_cluster_brokers
Total number of fully commissioned brokers configured in the cluster.
Type: gauge
Usage: Create an alert if this gauge falls below a steady-state threshold, which may indicate that a broker has become unresponsive.
redpanda_cluster_controller_log_limit_requests_available_rps
The upper limit on the requests per second (RPS) that the cluster controller log is allowed to process, segmented by command group.
Type: gauge
Labels:
-
redpanda_cmd_group=("move_operations" | "topic_operations" | "configuration_operations" | "node_management_operations" | "acls_and_users_operations")
redpanda_cluster_controller_log_limit_requests_dropped
The cumulative number of requests dropped by the controller log because the incoming rate exceeded the available RPS limit.
Type: counter
Labels:
-
redpanda_cmd_group=("move_operations" | "topic_operations" | "configuration_operations" | "node_management_operations" | "acls_and_users_operations")
Usage: A rising counter indicates that requests are being dropped, which could signal overload or misconfiguration.
redpanda_cluster_features_enterprise_license_expiry_sec
Number of seconds remaining until the Enterprise Edition license expires.
Type: gauge
Usage:
-
A value of
-1indicates that no license is present. -
A value of
0signifies an expired license.
Use this metric to proactively monitor license status and trigger alerts for timely renewal.
redpanda_cluster_members_backend_queued_node_operations
The number of node operations queued per shard that are awaiting processing by the backend.
Type: gauge
Labels:
-
shard
redpanda_cluster_members_backend_queued_node_operations
Number of queued node operations.
Type: gauge
Labels:
-
shard
redpanda_cluster_non_homogenous_fips_mode
Count of brokers whose FIPS mode configuration differs from the rest of the cluster.
Type: gauge
Usage: Indicates inconsistencies in security configurations that might affect compliance or interoperability.
redpanda_cluster_partition_moving_from_node
Number of partition replicas that are in the process of being removed from a broker.
Type: gauge
Usage: A non-zero value can indicate ongoing or unexpected partition reassignments. Investigate if this metric remains elevated.
redpanda_cluster_partition_moving_to_node
Number of partition replicas in the cluster currently being added or moved to a broker.
Type: gauge
Usage: When this gauge is non-zero, determine whether there is an expected or unexpected reassignment of partitions causing partition replicas movement.
redpanda_cluster_partition_node_cancelling_movements
During a partition movement cancellation operation, the number of partition replicas that were being moved that now need to be canceled.
Type: gauge
Usage: Track this metric to verify that partition reassignments are proceeding as expected; persistent non-zero values may warrant further investigation.
redpanda_cluster_partition_num_with_broken_rack_constraint
During a partition movement cancellation operation, the number of partition replicas that were scheduled for movement but now require cancellation.
Type: gauge
Usage: A non-zero value may indicate issues in the partition reassignment process that need attention.
redpanda_cluster_partitions
Total number of logical partitions managed by the cluster. This includes partitions for the controller topic but excludes replicas.
Type: gauge
redpanda_cluster_unavailable_partitions
Number of partitions that are unavailable due to a lack of quorum among their replica set.
Type: gauge
Usage: A non-zero value indicates that some partitions do not have an active leader. Consider increasing the number of brokers or the replication factor if this persists.
redpanda_cluster_latest_cluster_metadata_manifest_age
The amount of time in seconds since the last time Redpanda uploaded metadata files to Tiered Storage for your cluster. A value of 0 indicates metadata has not yet been uploaded.
When performing a whole cluster restore operation, the metadata for the new cluster will not have any changes made to a source cluster that is newer than this age.
Type: gauge
Usage: On a healthy system, this should not exceed the value set for cloud_storage_cluster_metadata_upload_interval_ms. You may consider setting an alert if this remains 0 for longer than 1.5 * cloud_storage_cluster_metadata_upload_interval_ms as that may indicate a configuration issue.
Debug bundle metrics
redpanda_debug_bundle_failed_generation_count
Running count of debug bundle generation failures, reported per shard.
Type: counter
Labels:
-
shard
redpanda_debug_bundle_last_failed_bundle_timestamp_seconds
Unix epoch timestamp of the last failed debug bundle generation, per shard.
Type: gauge
Labels:
-
shard
Infrastructure metrics
redpanda_cpu_busy_seconds_total
Total time (in seconds) the CPU has been actively processing tasks.
Type: counter
Usage: Useful for tracking overall CPU utilization.
redpanda_io_queue_total_read_ops
Cumulative count of read operations processed by the I/O queue.
Type: counter
Labels:
-
class=("default" | "compaction" | "raft") -
ioshard -
mountpoint -
shard
redpanda_io_queue_total_write_ops
Cumulative count of write operations processed by the I/O queue.
Type: counter
Labels:
-
class=("default" | "compaction" | "raft") -
ioshard -
mountpoint -
shard
redpanda_memory_allocated_memory
Total memory allocated (in bytes) per CPU shard.
Type: gauge
Labels:
-
shard
redpanda_memory_available_memory
Total memory (in bytes) available to a CPU shard—including both free and reclaimable memory.
Type: gauge
Labels:
-
shard
Usage: Indicates memory pressure on each shard.
redpanda_memory_available_memory_low_water_mark
The lowest recorded available memory (in bytes) per CPU shard since the process started.
Type: gauge
Labels:
-
shard
redpanda_memory_free_memory
Total unallocated (free) memory in bytes available for each CPU shard.
Type: gauge
Labels:
-
shard
redpanda_rpc_active_connections
Current number of active RPC client connections on a shard.
Type: gauge
Labels:
-
redpanda_server=("kafka" | "internal")
redpanda_rpc_request_errors_total
Cumulative count of RPC errors encountered, segmented by server type.
Type: counter
Labels:
-
redpanda_server=("kafka" | "internal")
Usage: Use this metric to diagnose potential issues in RPC communication.
redpanda_rpc_request_latency_seconds
Histogram capturing the latency (in seconds) for RPC requests.
Type: histogram
Labels:
-
redpanda_server=("kafka" | "internal")
redpanda_scheduler_runtime_seconds_total
Total accumulated runtime (in seconds) for the task queue associated with each scheduling group per shard.
Type: counter
Labels:
-
redpanda_scheduling_group=("admin" | "archival_upload" | "cache_background_reclaim" | "cluster" | "coproc" | "kafka" | "log_compaction" | "main" | "node_status" | "raft" | "raft_learner_recovery") -
shard
redpanda_storage_cache_disk_free_bytes
Amount of free disk space (in bytes) available on the cache storage.
Type: gauge
Usage: Monitor this to ensure sufficient cache storage capacity.
redpanda_storage_cache_disk_free_space_alert
Alert indicator for cache storage free space, where:
-
0= OK -
1= Low space -
2= Degraded
Type: gauge
redpanda_storage_disk_free_bytes
Amount of free disk space (in bytes) available on attached storage.
Type: gauge
Raft metrics
Service metrics
redpanda_authorization_result
Cumulative count of authorization results, categorized by result type.
Type: counter
Labels:
-
type
redpanda_kafka_rpc_sasl_session_expiration_total
Total number of SASL session expirations observed.
Type: counter
redpanda_kafka_rpc_sasl_session_reauth_attempts_total
Total number of SASL reauthentication attempts made by clients.
Type: counter
redpanda_kafka_rpc_sasl_session_revoked_total
Total number of SASL sessions that have been revoked.
Type: counter
redpanda_rest_proxy_request_latency_seconds
Histogram capturing the latency (in seconds) for REST proxy requests. The measurement includes waiting for resource availability, processing, and response dispatch.
Type: histogram
redpanda_schema_registry_cache_schema_count
Total number of schemas currently stored in the Schema Registry cache.
Type: gauge
redpanda_schema_registry_cache_schema_memory_bytes
Memory usage (in bytes) by schemas stored in the Schema Registry cache.
Type: gauge
redpanda_schema_registry_cache_subject_count
Count of subjects stored in the Schema Registry cache.
Type: gauge
Labels:
-
deleted
redpanda_schema_registry_cache_subject_version_count
Count of versions available for each subject in the Schema Registry cache.
Type: gauge
Labels:
-
deleted -
subject
redpanda_schema_registry_inflight_requests_memory_usage_ratio
Ratio of memory used by in-flight requests in the Schema Registry, reported per shard.
Type: gauge
Labels:
-
shard
redpanda_schema_registry_inflight_requests_usage_ratio
Usage ratio for in-flight Schema Registry requests, reported per shard.
Type: gauge
Labels:
-
shard
redpanda_schema_registry_queued_requests_memory_blocked
Count of Schema Registry requests queued due to memory constraints, reported per shard.
Type: gauge
Labels:
-
shard
Partition metrics
redpanda_kafka_max_offset
High watermark offset for a partition, used to calculate consumer group lag.
Type: gauge
Labels:
-
redpanda_namespace -
redpanda_partition -
redpanda_topic
Related topics:
redpanda_kafka_under_replicated_replicas
Number of partition replicas that are live yet lag behind the latest offset, redpanda_kafka_max_offset.
Type: gauge
Labels:
-
redpanda_namespace -
redpanda_partition -
redpanda_topic
redpanda_raft_leadership_changes
Total count of leadership changes (such as successful leader elections) across all partitions for a given topic.
Type: counter
Labels:
-
redpanda_namespace -
redpanda_topic
redpanda_raft_learners_gap_bytes
Total number of bytes that must be delivered to learner replicas to bring them up to date.
Type: gauge
Labels:
-
shard
redpanda_raft_recovery_offsets_pending
Sum of offsets across partitions on a broker that still need to be recovered.
Type: gauge
redpanda_raft_recovery_partition_movement_available_bandwidth
Available network bandwidth (in bytes per second) for partition movement operations.
Type: gauge
Labels:
-
shard
redpanda_raft_recovery_partition_movement_consumed_bandwidth
Network bandwidth (in bytes per second) currently being consumed for partition movement.
Type: gauge
Labels:
-
shard
Topic metrics
redpanda_cluster_partition_schema_id_validation_records_failed
Count of records that failed schema ID validation during ingestion.
Type: counter
redpanda_kafka_partitions
Configured number of partitions for a topic.
Type: gauge
Labels:
-
redpanda_namespace -
redpanda_topic
redpanda_kafka_replicas
Configured number of replicas for a topic.
Type: gauge
Labels:
-
redpanda_namespace -
redpanda_topic
redpanda_kafka_request_bytes_total
Total number of bytes processed (produced or consumed) for a topic.
Type: counter
Labels:
-
redpanda_namespace -
redpanda_topic -
redpanda_request=("produce" | "consume")
redpanda_raft_leadership_changes
Total number of leadership changes across partitions for a given topic.
Type: counter
Labels:
-
redpanda_namespace -
redpanda_topic
Broker metrics
redpanda_kafka_handler_latency_seconds
Histogram capturing the latency for processing Kafka requests at the broker level.
Type: histogram
redpanda_kafka_request_latency_seconds
Histogram capturing the latency (in seconds) for produce/consume requests at the broker. This duration spans from request initiation to response fulfillment.
Type: histogram
Labels:
-
redpanda_request=("produce" | "consume")
redpanda_kafka_quotas_client_quota_throttle_time
Histogram of client quota throttling delays (in seconds) per quota rule and type.
Type: histogram
Labels:
-
quota_rule=("not_applicable" | "kafka_client_default" | "cluster_client_default" | "kafka_client_prefix" | "cluster_client_prefix" | "kafka_client_id") -
quota_type=("produce_quota" | "fetch_quota" | "partition_mutation_quota")
redpanda_kafka_quotas_client_quota_throughput
Histogram of client quota throughput per quota rule and type.
Type: histogram
Labels:
-
quota_rule=("not_applicable" | "kafka_client_default" | "cluster_client_default" | "kafka_client_prefix" | "cluster_client_prefix" | "kafka_client_id") -
quota_type=("produce_quota" | "fetch_quota" | "partition_mutation_quota")
Consumer group metrics
REST proxy metrics
redpanda_rest_proxy_inflight_requests_memory_usage_ratio
Ratio of memory used by in-flight REST proxy requests, measured per shard.
Type: gauge
Labels:
-
shard
redpanda_rest_proxy_inflight_requests_usage_ratio
Usage ratio for in-flight REST proxy requests, measured per shard.
Type: gauge
Labels:
-
shard
redpanda_rest_proxy_queued_requests_memory_blocked
Count of REST proxy requests queued due to memory limitations, measured per shard.
Type: gauge
Labels:
-
shard
redpanda_rest_proxy_inflight_requests_memory_usage_ratio
Memory usage ratio of in-flight requests in the rest_proxy.
Type: gauge
Labels:
-
shard
redpanda_rest_proxy_inflight_requests_usage_ratio
Usage ratio of in-flight requests in the rest_proxy.
Type: gauge
Labels:
-
shard
redpanda_rest_proxy_queued_requests_memory_blocked
Number of requests queued in rest_proxy, due to memory limitations.
Type: gauge
Labels:
-
shard
Application metrics
redpanda_application_build
Build information for Redpanda, including the revision and version details.
Type: gauge
Labels:
-
redpanda_revision -
redpanda_version
Cloud metrics
redpanda_cloud_client_backoff
Total number of object storage requests that experienced backoff delays.
Type: counter
Labels:
-
For S3 and GCP:
-
redpanda_endpoint -
redpanda_region
-
-
For Azure Blob Storage (ABS):
-
redpanda_endpoint -
redpanda_storage_account
-
redpanda_cloud_client_client_pool_utilization
Current utilization (as a percentage, 0–100) of the object storage client pool.
Type: gauge
Labels:
-
redpanda_endpoint -
redpanda_region -
shard
redpanda_cloud_client_client_pool_utilization
Utilization of the object storage pool(0 - unused, 100 - fully utilized).
Type: gauge
Labels:
-
redpanda_endpoint -
redpanda_region -
shard
redpanda_cloud_client_download_backoff
Total number of object storage download requests that experienced backoff delays.
Type: counter
Labels:
-
For S3 and GCP:
-
redpanda_endpoint -
redpanda_region
-
-
For Azure Blob Storage (ABS):
-
redpanda_endpoint -
redpanda_storage_account
-
redpanda_cloud_client_downloads
Total number of successful download requests from object storage.
Type: counter
Labels:
-
For S3 and GCP:
-
redpanda_endpoint -
redpanda_region
-
-
For Azure Blob Storage (ABS):
-
redpanda_endpoint -
redpanda_storage_account
-
redpanda_cloud_client_lease_duration
Histogram representing the lease duration for object storage clients.
Type: histogram
redpanda_cloud_client_not_found
Total number of object storage requests that resulted in a “not found” error.
Type: counter
Labels:
-
For S3 and GCP:
-
redpanda_endpoint -
redpanda_region
-
-
For Azure Blob Storage (ABS):
-
redpanda_endpoint -
redpanda_storage_account
-
redpanda_cloud_client_num_borrows
Count of instances where a shard borrowed a object storage client from another shard.
Type: counter
Labels:
-
redpanda_endpoint -
redpanda_region -
shard
redpanda_cloud_client_num_borrows
Number of times the current shard had to borrow a object storage client from another shard.
Type: counter
Labels:
-
redpanda_endpoint -
redpanda_region -
shard
redpanda_cloud_client_upload_backoff
Total number of object storage upload requests that experienced backoff delays.
Type: counter
Labels:
-
For S3 and GCP:
-
redpanda_endpoint -
redpanda_region
-
-
For Azure Blob Storage (ABS):
-
redpanda_endpoint -
redpanda_storage_account
-
redpanda_cloud_client_uploads
Total number of successful upload requests to object storage.
Type: counter
Labels:
-
For S3 and GCP:
-
redpanda_endpoint -
redpanda_region
-
-
For Azure Blob Storage (ABS):
-
redpanda_endpoint -
redpanda_storage_account
-
redpanda_cloud_storage_cloud_log_size
Total size (in bytes) of user-visible log data stored in Tiered Storage. This value increases with every segment offload and decreases when segments are deleted due to retention or compaction.
Type: gauge
Usage: Segmented by redpanda_namespace (e.g., kafka, kafka_internal, or redpanda), redpanda_topic, and redpanda_partition.
TLS metrics
redpanda_tls_truststore_expires_at_timestamp_seconds
Unix epoch timestamp representing the expiration time of the shortest-lived certificate authority (CA) in the installed truststore.
Type: gauge
Labels:
-
area -
detail
Usage: Helps identify when any CA in the chain is nearing expiration.
redpanda_tls_certificate_expires_at_timestamp_seconds
Unix epoch timestamp for the expiration of the shortest-lived installed TLS certificate.
Type: gauge
Labels:
-
area -
detail
Usage: Useful for proactive certificate renewal by indicating the next certificate set to expire.
redpanda_tls_certificate_serial
The least significant 4 bytes of the serial number for the certificate that will expire next.
Type: gauge
Labels:
-
area -
detail
Usage: Provides a quick reference to identify the certificate in question.
Data transforms metrics
redpanda_transform_execution_latency_sec
Histogram tracking the execution latency (in seconds) for processing a single record via data transforms.
Type: histogram
Labels:
-
function_name
redpanda_transform_execution_errors
Counter for the number of errors encountered during the invocation of data transforms.
Type: counter
Labels:
-
function_name
redpanda_wasm_engine_cpu_seconds_total
Total CPU time (in seconds) consumed by WebAssembly functions.
Type: counter
Labels:
-
function_name
redpanda_wasm_engine_memory_usage
Current memory usage (in bytes) by a WebAssembly function.
Type: gauge
Labels:
-
function_name
redpanda_wasm_engine_max_memory
Maximum allowed memory (in bytes) allocated for a WebAssembly function.
Type: gauge
Labels:
-
function_name
redpanda_wasm_binary_executable_memory_usage
Amount of memory (in bytes) used by executable WebAssembly binaries.
Type: gauge
redpanda_transform_read_bytes
Cumulative count of bytes read as input to data transforms.
Type: counter
Labels:
-
function_name
redpanda_transform_write_bytes
Cumulative count of bytes output from data transforms.
Type: counter
Labels:
-
function_name
redpanda_transform_processor_lag
Number of records pending processing in the input topic for a data transform.
Type: gauge
Labels:
-
function_name
Object storage metrics
redpanda_cloud_storage_active_segments
Number of remote log segments that are currently hydrated and available for read operations.
Type: gauge
redpanda_cloud_storage_anomalies
Count of missing partition manifest anomalies detected for the topic.
Type: gauge
redpanda_cloud_storage_cache_op_hit
Total number of successful get requests that found the requested object in the cache.
Type: counter
redpanda_cloud_storage_cache_op_in_progress_files
Number of files currently being written to the cache.
Type: gauge
redpanda_cloud_storage_cache_op_miss
Total count of get requests that did not find the requested object in the cache.
Type: counter
redpanda_cloud_storage_cache_op_put
Total number of objects successfully written into the cache.
Type: counter
redpanda_cloud_storage_cache_space_hwm_files
High watermark for the number of objects stored in the cache.
Type: gauge
redpanda_cloud_storage_cache_space_hwm_size_bytes
High watermark for the total size (in bytes) of cached objects.
Type: gauge
redpanda_cloud_storage_cache_space_size_bytes
Total size (in bytes) of objects currently stored in the cache.
Type: gauge
redpanda_cloud_storage_cache_space_tracker_size
Current count of entries in the cache access tracker.
Type: gauge
redpanda_cloud_storage_cache_space_tracker_syncs
Total number of times the cache access tracker was synchronized with disk data.
Type: counter
redpanda_cloud_storage_cache_trim_carryover_trims
Count of times the cache trim operation was invoked using a carryover strategy.
Type: counter
redpanda_cloud_storage_cache_trim_exhaustive_trims
Count of instances where a fast cache trim was insufficient and an exhaustive trim was required.
Type: counter
redpanda_cloud_storage_cache_trim_failed_trims
Count of cache trim operations that failed to free the expected amount of space, possibly indicating a bug or misconfiguration.
Type: counter
redpanda_cloud_storage_cache_trim_fast_trims
Count of successful fast cache trim operations.
Type: counter
redpanda_cloud_storage_cache_trim_in_mem_trims
Count of cache trim operations performed using the in-memory access tracker.
Type: counter
redpanda_cloud_storage_cloud_log_size
Total size (in bytes) of the user-visible log data stored in Tiered Storage. This metric increases with every offloaded segment and decreases when segments are deleted due to retention or compaction.
Type: gauge
Usage: Labeled by redpanda_namespace (e.g., kafka, kafka_internal, or redpanda), redpanda_topic, and redpanda_partition.
redpanda_cloud_storage_deleted_segments
Count of log segments that have been deleted from object storage due to retention policies or compaction processes.
Type: counter
redpanda_cloud_storage_errors_total
Cumulative count of errors encountered during object storage operations, segmented by direction.
Type: counter
Labels:
-
redpanda_direction
redpanda_cloud_storage_housekeeping_drains
Count of times the object storage upload housekeeping queue was fully drained.
Type: gauge
redpanda_cloud_storage_housekeeping_jobs_completed
Total number of successfully executed object storage housekeeping jobs.
Type: counter
redpanda_cloud_storage_housekeeping_jobs_failed
Total number of object storage housekeeping jobs that failed.
Type: counter
redpanda_cloud_storage_housekeeping_jobs_skipped
Count of object storage housekeeping jobs that were skipped during execution.
Type: counter
redpanda_cloud_storage_housekeeping_pauses
Count of times object storage upload housekeeping was paused.
Type: gauge
redpanda_cloud_storage_housekeeping_requests_throttled_average_rate
Average rate (per shard) of requests that were throttled during object storage operations.
Type: gauge
redpanda_cloud_storage_housekeeping_resumes
Count of instances when object storage upload housekeeping resumed after a pause.
Type: gauge
redpanda_cloud_storage_housekeeping_rounds
Total number of rounds executed by the object storage upload housekeeping process.
Type: counter
redpanda_cloud_storage_jobs_cloud_segment_reuploads
Count of log segments reuploaded from object storage sources (either from the cache or via direct download).
Type: gauge
redpanda_cloud_storage_jobs_local_segment_reuploads
Count of log segments reuploaded from the local data directory.
Type: gauge
redpanda_cloud_storage_jobs_manifest_reuploads
Total number of partition manifest reuploads performed by housekeeping jobs.
Type: gauge
redpanda_cloud_storage_jobs_metadata_syncs
Total number of archival configuration updates (metadata synchronizations) executed by housekeeping jobs.
Type: gauge
redpanda_cloud_storage_jobs_segment_deletions
Total count of log segments deleted by housekeeping jobs.
Type: gauge
redpanda_cloud_storage_limits_downloads_throttled_sum
Total cumulative time (in milliseconds) during which downloads were throttled.
Type: counter
redpanda_cloud_storage_partition_manifest_uploads_total
Total number of successful partition manifest uploads to object storage.
Type: counter
redpanda_cloud_storage_partition_readers
Number of active partition reader instances (fetch/timequery operations) reading from Tiered Storage.
Type: gauge
redpanda_cloud_storage_partition_readers_delayed
Count of partition read operations delayed due to reaching the reader limit, suggesting potential saturation of Tiered Storage reads.
Type: counter
redpanda_cloud_storage_readers
Total number of segment read cursors for hydrated remote log segments.
Type: gauge
redpanda_cloud_storage_segment_index_uploads_total
Total number of successful segment index uploads to object storage.
Type: counter
redpanda_cloud_storage_segment_materializations_delayed
Count of segment materialization operations that were delayed because of reader limit constraints.
Type: counter
redpanda_cloud_storage_segment_readers_delayed
Count of segment reader operations delayed due to reaching the reader limit. This indicates a cluster is saturated with Tiered Storage reads.
Type: counter
redpanda_cloud_storage_segment_uploads_total
Total number of successful data segment uploads to object storage.
Type: counter
redpanda_cloud_storage_segments
Total number of log segments accounted for in object storage for the topic.
Type: gauge
redpanda_cloud_storage_segments_pending_deletion
Count of log segments pending deletion from object storage for the topic.
Type: gauge
redpanda_cloud_storage_spillover_manifest_uploads_total
Total number of successful spillover manifest uploads to object storage.
Type: counter
redpanda_cloud_storage_spillover_manifests_materialized_bytes
Total bytes of memory used by spilled manifests that are currently cached in memory.
Type: gauge
