Monitoring metrics

This document describes the metrics provided by the KakaoCloud Monitoring service.

Virtual Machine, GPU, and Bare Metal Server metrics

These are the primary system resource metrics collected commonly across Virtual Machines, GPUs, and Bare Metal Servers. They can be utilized in the following service areas:

• Monitoring: Custom dashboards, Metric Explorer, Metric Export
• Alert Center: Setting metric-based notification policies

CPU and memory

Metric name	Description	Unit	Recommended use
cpu_usage	Total CPU utilization	%	Key performance indicator
cpu_usage_user	CPU utilization (User processes)	%	Check user process load
cpu_usage_system	CPU utilization (System kernel)	%	Check kernel/system load
cpu_usage_iowait	CPU utilization (I/O wait)	%	Diagnose I/O bottlenecks
cpu_usage_per_core	CPU utilization per core	%	Check load imbalance across cores
mem_usage	Self-memory utilization	%	Major memory alerts
mem_used	Used memory size	bytes(IEC)	Check absolute usage
mem_buffered	Memory usage (Buffers)	bytes(IEC)	Linux only
mem_cached	Memory usage (Cache)	bytes(IEC)	Linux only

Disk I/O and capacity

Metric name	Description	Unit	Recommended use
disk_used_percent	Disk utilization	%	Capacity shortage notification
disk_used	Disk usage	bytes(IEC)	Check absolute usage
disk_read_bytes_persec	Disk read bytes per second	bytes/s(IEC)	Check read performance latency
disk_write_bytes_persec	Disk write bytes per second	bytes/s(IEC)	Check write performance latency
disk_read_iops	Disk read IOPS	count/s	Check read operation load
disk_write_iops	Disk write IOPS	count/s	Check write operation load
disk_inodes_usage	Disk inode utilization	%	inode shortage alert (Linux)
disk_free	Available disk capacity	bytes(IEC)	Reference information
disk_total	Total disk capacity	bytes(IEC)	Reference information
disk_inodes_free	Available inode count	count	Reference information
disk_inodes_total	Total reserved inode count	count	Reference information
disk_inodes_used	inode usage	count	Reference information

Network

Metric name	Description	Unit	Recommended use
network_rx_bytes_persec	Bytes received per second on network interface	bytes/s(IEC)	Check inbound traffic
network_tx_bytes_persec	Bytes sent per second on network interface	bytes/s(IEC)	Check outbound traffic
network_rx_packets_persec	Packets received per second on network interface	packets/s	Check inbound packet load
network_tx_packets_persec	Packets sent per second on network interface	packets/s	Check outbound packet load

GPU-specific metrics

Metric name	Description	Unit	Recommended use
nvidia_smi_utilization_gpu	GPU utilization per core	%	Check GPU performance load
nvidia_smi_memory_used	Memory used per GPU core	MiB(IEC)	Check GPU memory usage
nvidia_smi_memory_free	Available memory per GPU core	MiB(IEC)	Check GPU memory margin
nvidia_smi_memory_total	Total memory per GPU core	MiB(IEC)	Check total GPU memory
nvidia_smi_power_draw	Power consumption per GPU core	watt	Check power usage

Virtual Machine metrics for burstable instances

The following metrics are collected only for t1i family instances with the burstable option applied (excluding the t1i.medium.dns.default type).

Metric name	Description	Unit
cpu_credit_usage	Accumulated CPU credit usage; credits consumed when CPU usage exceeds baseline performance	count
cpu_credit_balance	Remaining CPU credit balance; earned when operating below baseline performance	count

info

• mem_buffered, mem_cached, and disk_inodes_usage metrics are only collected and provided for servers with Linux OS installed.
• nvidia_smi metrics are only collected for servers equipped with GPUs.

caution

• GPU instance library compatibility: When updating NVIDIA libraries for GPU instances, ensure you check CUDA version compatibility. If incompatible, the monitoring agent may fail to collect NVIDIA metrics.
• Network notification policy: When setting an Alert Center notification policy using the network_rx_bytes_persec metric, the policy applies to all network interfaces. In multi-NIC instances, an alert will be sent if any connected interface exceeds the threshold.

Libvirt metrics

These are the primary resource metrics for virtualization-based servers collected in Libvirt environments. They can be used in:

• Monitoring: Custom dashboards, Metric Explorer, Metric Export

CPU and memory

Metric name	Description	Unit
libvirt_domain_info_cpu_time_seconds_total	Total CPU time used	count
libvirt_domain_info_virtual_cpus	Number of CPU cores	count

Disk I/O and capacity

Metric name	Description	Unit
libvirt_domain_block_stats_read_bytes_total	Total bytes read from disk	bytes(IEC)
libvirt_domain_block_stats_write_bytes_total	Total bytes written to disk	bytes(IEC)

Network

Metric name	Description	Unit
libvirt_domain_interface_stats_receive_bytes_total	Total bytes received on network interface	bytes(IEC)
libvirt_domain_interface_stats_transmit_bytes_total	Total bytes sent on network interface	bytes(IEC)
libvirt_domain_interface_stats_receive_packets_total	Total packets received on network interface	packets
libvirt_domain_interface_stats_transmit_packets_total	Total packets sent on network interface	packets
libvirt_domain_interface_stats_receive_drops_total	Total packets dropped on reception	packets

Kubernetes Engine metrics

Primary cluster, node, and pod resource metrics collected in Kubernetes Engine environments. They can be used in:

• Monitoring: Metric Export

Cluster

Metric name	Description	Unit
cluster_autoscaler_node_group_min_count	Minimum node count for node group autoscaling	count
cluster_autoscaler_node_group_max_count	Maximum node count for node group autoscaling	count
cluster_autoscaler_node_group_target_count	Target node count for node group autoscaling	count
node_count	Current number of nodes	count

Node

Metric name	Description	Unit
kube_node_status_allocatable	Resource amount allocatable to pods on the node	none
kube_node_status_capacity	Total resource capacity of the node	none
node_cpu_seconds_total	Time used by node CPU in each mode	count
node_filesystem_avail_bytes	Filesystem capacity available to non-root users on the node	bytes(IEC)
node_filesystem_size_bytes	Total size of the node filesystem	bytes(IEC)
node_memory_Active_bytes	Memory currently in use or reclaimable (Active)	bytes(IEC)
node_memory_Buffers_bytes	Memory used as kernel disk I/O buffers	bytes(IEC)
node_memory_Cached_bytes	Memory used as filesystem cache	bytes(IEC)
node_memory_MemAvailable_bytes	Memory immediately available for new processes	bytes(IEC)
node_memory_MemFree_bytes	Memory currently unallocated	bytes(IEC)
node_memory_MemTotal_bytes	Total memory capacity of the node	bytes(IEC)
node_memory_SReclaimable_bytes	Reclaimable part of the Slab cache	bytes(IEC)
node_network_receive_bytes_total	Total bytes received by network devices	bytes(IEC)
node_network_transmit_bytes_total	Total bytes sent by network devices	bytes(IEC)

Pod and container

Metric name	Description	Unit
kube_pod_container_info	Basic information about containers within a pod	none
kube_pod_container_resource_limits	Resource limit requested by the container	none
kube_pod_container_resource_requests	Resource request value for the container	none
kube_pod_container_status_running	Whether the container status is Running	count
kube_pod_container_status_terminated	Whether the container status is Terminated	count
kube_pod_info	Information about the pod	none
container_cpu_usage_seconds_total	Total CPU time consumed by the container	count
container_memory_working_set_bytes	Memory in use by the container that the OS cannot reclaim	bytes(IEC)
container_network_receive_bytes_total	Total network bytes received by the container	bytes(IEC)
container_network_transmit_bytes_total	Total network bytes sent by the container	bytes(IEC)

Load Balancing metrics

Primary metrics for monitoring traffic and connection status of Load Balancer resources. Used in:

• Monitoring: Custom dashboards, Metric Explorer, Metric Export
• Alert Center: Setting metric-based notification policies

Metric name	Description	Unit
lb_bytes_in_persec	Inbound traffic per second (Received bytes)	bytes/s(IEC)
lb_bytes_out_persec	Outbound traffic per second (Sent bytes)	bytes/s(IEC)
lb_connections_persec	Connections created per second	count/s
lb_current_connections	Number of currently maintained connections	count
lb_healthy_host_count	Number of healthy hosts available for connection	count
lb_unhealthy_host_count	Number of unhealthy hosts unavailable for connection	count

Transit Gateway metrics

Primary metrics for monitoring traffic and connection status of Transit Gateway resources. These can be used in:

• Monitoring: Metric Export

Metric name	Description	Unit
tgw_packet_bytes	Total data size processed by the Transit Gateway	Bytes(IEC)
tgw_packet_count	Total number of packets processed by the Transit Gateway	Count

MySQL metrics

Primary metrics for monitoring storage, network, query, and connection status of MySQL instances. Used in:

• Monitoring: Custom dashboards, Metric Explorer, Metric Export
• Alert Center: Setting metric-based notification policies

CPU and memory

Metric name	Description	Unit	Note
cpu_usage	Total CPU utilization	%	Key performance indicator
cpu_usage_user	CPU utilization (User processes)	%
cpu_usage_system	CPU utilization (System kernel)	%
cpu_usage_iowait	CPU utilization (I/O wait)	%
cpu_usage_per_core	CPU utilization per core	%
mem_swap_total	Total swap memory	bytes(IEC)
mem_swap_cached	Cached swap memory	bytes(IEC)
mem_swap_free	Available swap memory	bytes(IEC)
mem_usage	Self-memory utilization	%
mem_used	Used memory size	bytes(IEC)
mem_buffered	Memory usage (Buffers)	bytes(IEC)
mem_cached	Memory usage (Cache)	bytes(IEC)

Disk I/O and capacity

Metric name	Description	Unit
mysql_logstorage_disk_write_bytes_persec	Bytes written per second to log storage disk	bytes/s(IEC)
mysql_defaultstorage_disk_write_bytes_persec	Bytes written per second to default storage disk	bytes/s(IEC)
mysql_logstorage_disk_read_bytes_persec	Bytes read per second from log storage disk	bytes/s(IEC)
mysql_defaultstorage_disk_read_bytes_persec	Bytes read per second from default storage disk	bytes/s(IEC)
mysql_logstorage_disk_write_iops	Write IOPS on log storage disk	count/s
mysql_defaultstorage_disk_write_iops	Write IOPS on default storage disk	count/s
mysql_logstorage_disk_read_iops	Read IOPS on log storage disk	count/s
mysql_defaultstorage_disk_read_iops	Read IOPS on default storage disk	count/s
mysql_logstorage_disk_used	Log storage disk usage	bytes(IEC)
mysql_defaultstorage_disk_used	Default storage disk usage	bytes(IEC)
mysql_logstorage_disk_used_percent	Log storage disk utilization	%
mysql_defaultstorage_disk_used_percent	Default storage disk utilization	%
mysql_logstorage_disk_inodes_usage	Log storage inode utilization	%
mysql_defaultstorage_disk_inodes_usage	Default storage inode utilization	%
mysql_defaultstorage_disk_free	Available capacity on default storage disk	bytes(IEC)
mysql_defaultstorage_disk_total	Total capacity of default storage disk	bytes(IEC)
mysql_logstorage_disk_free	Available capacity on log storage disk	bytes(IEC)
mysql_logstorage_disk_total	Total capacity of log storage disk	bytes(IEC)
mysql_defaultstorage_disk_inodes_free	Available inode count on default storage disk	count
mysql_defaultstorage_disk_inodes_total	Total inode count of default storage disk	count
mysql_defaultstorage_disk_inodes_used	inode usage on default storage disk	count
mysql_logstorage_disk_inodes_free	Available inode count on log storage disk	count
mysql_logstorage_disk_inodes_total	Total inode count of log storage disk	count
mysql_logstorage_disk_inodes_used	inode usage on log storage disk	count

Network

Metric name	Description	Unit
mysql_network_rx_bytes_persec	Bytes received per second on network interface	bytes/s(IEC)
mysql_network_tx_bytes_persec	Bytes sent per second on network interface	bytes/s(IEC)
mysql_network_rx_packets_persec	Packets received per second on network interface	packets/s
mysql_network_tx_packets_persec	Packets sent per second on network interface	packets/s

Query

Metric name	Description	Unit
mysql_query_persec	Queries per second (QPS)	count/s
mysql_com_insert_count	INSERT query count over 5 minutes	count
mysql_com_select_count	SELECT query count over 5 minutes	count
mysql_com_update_count	UPDATE query count over 5 minutes	count
mysql_com_delete_count	DELETE query count over 5 minutes	count
mysql_com_commit_count	COMMIT query count over 5 minutes	count
mysql_slow_query_count	Slow query count over 5 minutes	count

Connection and instance

Metric name	Description	Unit
mysql_connections_count	Number of currently connected sessions	count
mysql_max_connections_count	Maximum number of allowable connections	count
mysql_connection_usage_percent	Ratio of current connections to maximum allowable	%
mysql_instance_status	Instance status	count
mysql_instance_group_status	Instance group status	count
mysql_uptime	Instance uptime	duration

InnoDB and binary log

Metric name	Description	Unit
mysql_innodb_buffer_pool_read_requests	Total requests to the Buffer Pool	count
mysql_innodb_buffer_pool_reads	Requests read directly from the Buffer Pool	count
mysql_innodb_buffer_cache_hit_ratio	Buffer Pool cache hit ratio	%
mysql_innodb_row_lock_current_waits	Number of current row locks	count
mysql_innodb_row_lock_time	Time spent on row locks	milliseconds
mysql_binary_size_bytes	Binary log size	bytes(IEC)
mysql_binary_files_count	Number of binary log files	count
mysql_variables_max_binlog_size	Maximum binary log size setting	bytes(IEC)
mysql_replication_lag	Binary log replication delay	seconds

PostgreSQL metrics

Primary metrics for monitoring the disk, network, connection, and transaction status of PostgreSQL instances. These can be used in the following service areas:

• Monitoring: Custom dashboards, Metric Explorer, Metric Export
• Alert Center: Setting metric-based notification policies

CPU and memory

Metric name	Description	Unit	Note
cpu_usage	Total CPU utilization	%	Key performance indicator
cpu_usage_user	CPU utilization (User processes)	%
cpu_usage_system	CPU utilization (System kernel)	%
cpu_usage_iowait	CPU utilization (I/O wait)	%
cpu_usage_per_core	CPU utilization per core	%
mem_swap_total	Total swap memory	bytes(IEC)
mem_swap_cached	Cached swap memory	bytes(IEC)
mem_swap_free	Available swap memory	bytes(IEC)
mem_usage	Self-memory utilization	%
mem_used	Used memory size	bytes(IEC)
mem_buffered	Memory usage (Buffers)	bytes(IEC)
mem_cached	Memory usage (Cache)	bytes(IEC)

Disk I/O and capacity

Metric name	Description	Unit
pg_defaultstorage_disk_read_bytes_persec	Bytes read per second from default storage disk	bytes/s(IEC)
pg_defaultstorage_disk_write_bytes_persec	Bytes written per second to default storage disk	bytes/s(IEC)
pg_defaultstorage_disk_read_iops	Number of read operations completed per second on default storage disk	count/s
pg_defaultstorage_disk_write_iops	Number of write operations completed per second on default storage disk	count/s
pg_defaultstorage_disk_used	Default storage disk usage	bytes(IEC)
pg_defaultstorage_disk_used_percent	Default storage disk utilization	%
pg_defaultstorage_disk_inodes_usage	Default storage inode utilization	%
pg_defaultstorage_disk_free	Available capacity on default storage disk	bytes(IEC)
pg_defaultstorage_disk_total	Total capacity of default storage disk	bytes(IEC)
pg_defaultstorage_disk_inodes_free	Available inode count on default storage disk	count
pg_defaultstorage_disk_inodes_total	Total reserved inode count on default storage disk	count
pg_defaultstorage_disk_inodes_used	inode usage on default storage disk	count
pg_logstorage_disk_read_bytes_persec	Bytes read per second from log storage disk	bytes/s(IEC)
pg_logstorage_disk_write_bytes_persec	Bytes written per second to log storage disk	bytes/s(IEC)
pg_logstorage_disk_read_iops	Number of read operations completed per second on log storage disk	count/s
pg_logstorage_disk_write_iops	Number of write operations completed per second on log storage disk	count/s
pg_logstorage_disk_used	Log storage disk usage	bytes(IEC)
pg_logstorage_disk_used_percent	Log storage disk utilization	%
pg_logstorage_disk_inodes_usage	Log storage inode utilization	%
pg_logstorage_disk_free	Available capacity on log storage disk	bytes(IEC)
pg_logstorage_disk_total	Total capacity of log storage disk	bytes(IEC)
pg_logstorage_disk_inodes_free	Available inode count on log storage disk	count
pg_logstorage_disk_inodes_total	Total reserved inode count on log storage disk	count
pg_logstorage_disk_inodes_used	inode usage on log storage disk	count

Network

Metric name	Description	Unit
pg_network_rx_bytes_persec	Bytes received per second on network interface	bytes/s(IEC)
pg_network_tx_bytes_persec	Bytes sent per second on network interface	bytes/s(IEC)
pg_network_rx_packets_persec	Packets received per second on network interface	packets/s
pg_network_tx_packets_persec	Packets sent per second on network interface	packets/s

Connection and operational status

Metric name	Description	Unit
pg_active_time	Total time consumed by active sessions	ms
pg_idle_in_transaction_time	Duration of "Idle in transaction" state	ms
pg_numbackends	Number of currently connected backend processes	count
pg_session_time	Duration sessions remained in an active state	ms
pg_sessions	Total number of sessions	count
pg_sessions_abandoned	Number of forced session terminations (Disconnected)	count
pg_sessions_fatal	Number of sessions terminated due to fatal errors	count
pg_sessions_killed	Number of sessions forcibly terminated by an operator	count
pg_datid	Database identifier (ID)	count

Replication and transactions

Metric name	Description	Unit
pg_replication_lag	Replication delay time	seconds
pg_xact_commit	Number of committed transactions	count
pg_xact_rollback	Number of rolled-back transactions	count
pg_xid_age	XID age of a specific process	count
pg_xid_age_per_group	XID age per instance group	count
pg_temp_file_ratio	Ratio of temporary files to total transactions	%
pg_temp_file_ratio_per_group	Ratio of temporary files per instance group	%

Workload and performance

Metric name	Description	Unit
pg_blk_read_time	Block read time	ms
pg_blk_write_time	Block write time	ms
pg_blks_hit	Number of blocks read from cache	count
pg_blks_read	Number of blocks read from disk	count
pg_buffer_hit_ratio	Buffer hit ratio	%
pg_checksum_failures	Number of checksum failures	count
pg_conflicts	Number of transaction conflicts	count
pg_deadlocks	Number of deadlocks occurred	count
pg_temp_bytes	Temporary file usage	bytes
pg_temp_files	Number of temporary files created	count
pg_tup_deleted	Number of deleted tuples	count
pg_tup_fetched	Number of fetched tuples	count
pg_tup_inserted	Number of inserted tuples	count
pg_tup_returned	Number of returned tuples	count
pg_tup_updated	Number of updated tuples	count

MemStore metrics

Primary metrics for monitoring the memory, network, replication, and CPU usage status of MemStore instances. These can be used in the following service areas:

• Monitoring: Custom dashboards, Metric Explorer, Metric Export
• Alert Center: Setting metric-based notification policies

CPU and memory

Metric name	Description	Unit
memstore_used_cpu_sys	Total system CPU usage	count
memstore_used_cpu_sys_main_thread	System CPU usage of the main thread	count
memstore_used_cpu_user	Total user CPU usage	count
memstore_used_cpu_user_main_thread	User CPU usage of the main thread	count
memstore_memory_usage	Total memory utilization	%
memstore_used_memory	Memory size currently in use by MemStore	bytes(IEC)
memstore_used_memory_peak	Peak memory used	bytes(IEC)
memstore_used_memory_peak_perc	Ratio of peak usage to total memory	%
memstore_used_memory_dataset	Memory used for actual data storage	bytes(IEC)
memstore_used_memory_dataset_perc	Ratio of memory used for actual data storage	%
memstore_used_memory_overhead	Overhead memory required for internal data structure management	bytes(IEC)
memstore_used_memory_lua	Memory used for Lua script execution	bytes(IEC)
memstore_allocator_allocated	Memory allocated to the allocator (including internal fragmentation)	bytes(IEC)
memstore_allocator_active	Active memory in the allocator (including external fragmentation)	bytes(IEC)
memstore_allocator_resident	Resident memory managed by the allocator	bytes(IEC)
memstore_allocator_rss_bytes	RSS memory size	bytes(IEC)
memstore_allocator_frag_bytes	Difference between active memory and allocated memory	bytes(IEC)
memstore_allocator_frag_ratio	Ratio of allocated memory to active memory	%
memstore_allocator_rss_ratio	Ratio of active memory to resident memory	%
memstore_mem_fragmentation_bytes	Difference between resident memory in use and allocated memory	bytes(IEC)
memstore_mem_fragmentation_ratio	Ratio of resident memory in use to allocated memory	%
memstore_rss_overhead_bytes	Difference between process RSS and allocator resident memory	bytes(IEC)
memstore_rss_overhead_ratio	Ratio of process RSS to allocator resident memory	%
memstore_total_system_memory	Total memory of the system where MemStore is running	bytes(IEC)

Disk I/O and capacity

Metric name	Description	Unit
disk_free	Available disk capacity	bytes(IEC)
disk_total	Total disk capacity	bytes(IEC)
disk_inodes_free	Available inode count	count
disk_inodes_used	Used inode count	count
disk_inodes_total	Total reserved inode count	count

Network

Metric name	Description	Unit
memstore_instantaneous_input_kbps	Network input rate per second	KiB/s(IEC)
memstore_instantaneous_output_kbps	Network output rate per second	KiB/s(IEC)
memstore_total_net_input_bytes	Total network input bytes	bytes(IEC)
memstore_total_net_output_bytes	Total network output bytes	bytes(IEC)
memstore_instantaneous_ops_per_sec	Number of commands processed per second	count
memstore_cmdstat_calls_persec	Number of command calls per second	count/s
memstore_total_commands_processed	Total number of commands processed	count
memstore_total_reads_processed	Total number of read events processed	count
memstore_total_writes_processed	Total number of write events processed	count
memstore_io_threaded_reads_processed	Number of read events processed in I/O threads	count
memstore_io_threaded_writes_processed	Number of write events processed in I/O threads	count

Connection and operational status

Metric name	Description	Unit
memstore_connected_slaves	Number of connected replicas	count
memstore_replication_lag	Replica replication delay time	s
memstore_clients	Number of currently connected clients	count
memstore_maxclients	Maximum number of allowable client connections	count
memstore_client_ratio	Ratio of current client connections to the maximum	%
memstore_blocked_clients	Number of waiting clients (e.g., BLPOP commands)	count
memstore_cluster_connections	Number of sockets in use on the cluster bus	count
memstore_cluster_enabled	Whether the cluster is enabled	count
memstore_pubsub_channels	Number of Pub/Sub channels	count
memstore_pubsub_patterns	Number of Pub/Sub patterns	count
memstore_uptime	Instance uptime	s

Key and cache statistics

Metric name	Description	Unit
memstore_keyspace_hits	Number of key hits	count
memstore_keyspace_misses	Number of key misses	count
memstore_keyspace_hitrate_percent	Key hit rate	%
memstore_evicted_keys	Number of keys evicted due to memory limits	count
memstore_expired_keys	Number of expired keys	count
memstore_lazyfree_pending_objects	Number of objects pending release via Lazy Free	count
memstore_lazyfreed_objects	Number of objects released via Lazy Free	count
memstore_lru_clock	Internal time value for the LRU algorithm	count

Metrics for burstable instances

The following metrics are collected only for t1i family instances with the burstable option applied.

Metric name	Description	Unit
cpu_credit_usage	CPU credit usage	count
cpu_credit_balance	Remaining CPU credit balance	count

Hadoop Eco metrics

Primary HBase, HDFS, Yarn, and Kafka related system metrics collected in a Hadoop Eco environment. These can be used in:

• Monitoring: Metric Export

HBase

Metric name	Description	Unit
HBase_Master_JvmMetrics_MemHeapMaxM	Maximum JVM heap memory size for HBase Master	MB
HBase_Master_JvmMetrics_MemHeapUsedM	JVM heap memory usage for HBase Master	MB
HBase_Master_Server_numDeadRegionServers	Number of Region Servers in an abnormal (Dead) state	count
HBase_Master_Server_numRegionServers	Number of Region Servers operating normally	count

HDFS (NameNode)

Metric name	Description	Unit
Hadoop_NameNode_JvmMetrics_MemHeapMaxM	Maximum JVM heap memory size for NameNode	MB
Hadoop_NameNode_JvmMetrics_MemHeapUsedM	JVM heap memory usage for NameNode	MB
Hadoop_NameNode_JvmMetrics_GcTimeMillis	GC execution time for NameNode JVM	count
Hadoop_NameNode_FSNamesystem_CapacityTotal	Total HDFS storage capacity	bytes(IEC)
Hadoop_NameNode_FSNamesystem_CapacityUsed	HDFS storage capacity in use	bytes(IEC)
Hadoop_NameNode_FSNamesystem_CapacityRemaining	Available HDFS storage capacity	bytes(IEC)
Hadoop_NameNode_FSNamesystem_CapacityUsedNonDFS	Capacity used for non-DFS purposes (logs, temporary files, etc.)	bytes(IEC)
Hadoop_NameNode_FSNamesystem_NumLiveDataNodes	Number of DataNodes operating normally	count
Hadoop_NameNode_FSNamesystem_NumDeadDataNodes	Number of DataNodes in an abnormal (Dead) state	count
Hadoop_NameNode_FSNamesystem_StaleDataNodes	Number of DataNodes with suspended status updates	count
Hadoop_NameNode_FSNamesystem_CorruptBlocks	Number of corrupted blocks	count
Hadoop_NameNode_FSNamesystem_TotalLoad	Current number of active client connections	count

Yarn (ResourceManager)

Metric name	Description	Unit
Yarn_ResourceManager_JvmMetrics_MemHeapMaxM	Maximum JVM heap memory size for ResourceManager	MB
Yarn_ResourceManager_JvmMetrics_MemHeapUsedM	JVM heap memory usage for ResourceManager	MB
Yarn_ResourceManager_JvmMetrics_GcTimeMillis	GC execution time for ResourceManager JVM	count
Yarn_ResourceManager_ClusterMetrics_NumActiveNMs	Number of NodeManagers operating normally	count
Yarn_ResourceManager_ClusterMetrics_NumShutdownNMs	Number of terminated NodeManagers	count
Yarn_ResourceManager_QueueMetrics_AllocatedMB	Allocated memory size	MB
Yarn_ResourceManager_QueueMetrics_AvailableMB	Available memory size	MB
Yarn_ResourceManager_QueueMetrics_PendingMB	Pending memory size	MB
Yarn_ResourceManager_QueueMetrics_ReservedMB	Reserved memory size	MB
Yarn_ResourceManager_QueueMetrics_AllocatedVCores	Number of allocated vCores	count
Yarn_ResourceManager_QueueMetrics_AvailableVCores	Number of available vCores	count
Yarn_ResourceManager_QueueMetrics_PendingVCores	Number of pending vCores	count
Yarn_ResourceManager_QueueMetrics_ReservedVCores	Number of reserved vCores	count
Yarn_ResourceManager_QueueMetrics_AppsRunning	Number of running applications	count
Yarn_ResourceManager_QueueMetrics_AppsCompleted	Number of completed applications	count
Yarn_ResourceManager_QueueMetrics_AppsFailed	Number of failed applications	count
Yarn_ResourceManager_QueueMetrics_AppsKilled	Number of forcibly terminated applications	count
Yarn_ResourceManager_QueueMetrics_AppsPending	Number of pending applications	count
Yarn_ResourceManager_QueueMetrics_AppsSubmitted	Number of submitted applications	count

Kafka

Metric name	Description	Unit
Kafka_Active_Brokers	Number of Brokers operating normally	count
Kafka_Total_Topics	Number of topics currently in operation	count

Pub/Sub metrics

Primary metrics for monitoring the status of message publishing, subscription, and storage for the Pub/Sub service. These can be used in:

• Monitoring: Custom dashboards, Metric Explorer
• Alert Center: Setting metric-based notification policies

Publish

Metric name	Description	Unit
pubsub_published_message_count_persec	Number of published messages per second	count/s
pubsub_published_message_bytes_persec	Size of published messages per second	bytes/s(IEC)
pubsub_publish_request_count_persec	Number of publish requests per second	count/s
pubsub_topic_storage_used_bytes	Size of stored data in topic	bytes(IEC)

Subscription

Metric name	Description	Unit
pubsub_ack_request_count_persec	Number of acknowledgment (Ack) requests per second	count/s
pubsub_acked_message_count_persec	Number of acknowledged messages per second	count/s
pubsub_unprocessed_messages	Number of unprocessed messages	count
pubsub_pulled_message_count_persec	Number of pulled messages per second	count/s
pubsub_streaming_pull_response_count_persec	Number of streaming pull responses per second	count/s
pubsub_push_count_persec	Number of push requests per second	count/s
pubsub_pushed_message_count_persec	Number of pushed messages per second	count/s
pubsub_subscription_storage_used_bytes	Size of stored data in subscription	bytes(IEC)
pubsub_seek_request_count_permin	Number of Seek requests per minute	count/m

Storage and export

Metric name	Description	Unit
pubsub_exported_message_count_persec	Number of messages exported to Object Storage per second	count/s
pubsub_object_storage_api_call_count_permin	Number of Object Storage API calls per minute	count/m

Advanced Managed Kafka metrics

Primary metrics for monitoring brokers, partitions, topic throughput, and server resource status of Advanced Managed Kafka instances. These can be used in:

• Monitoring: Metric Export
• Alert Center: Setting metric-based notification policies

CPU and memory

Metric name	Description	Unit
cpu_usage	Overall CPU usage	%
mem_usage	Memory usage	%

Disk I/O and capacity

Metric name	Description	Unit
disk_used_percent	Disk usage	%
disk_read_bytes_persec	Amount of data read from disk per second	bytes/s(IEC)
disk_write_bytes_persec	Amount of data written to disk per second	bytes/s(IEC)
disk_read_iops	Number of disk read operations completed per second	count/s
disk_write_iops	Number of disk write operations completed per second	count/s

Network

Metric name	Description	Unit
network_rx_bytes_persec	Inbound traffic on the network interface	bytes/s(IEC)
network_tx_bytes_persec	Outbound traffic on the network interface	bytes/s(IEC)
network_rx_packets_persec	Number of packets received per second on the network interface	packets/s
network_tx_packets_persec	Number of packets sent per second on the network interface	packets/s

Cluster and partitions

Metric name	Description	Unit
kafka_controller_ActiveBrokerCount_Value	Number of active brokers in the cluster	count
kafka_controller_GlobalPartitionCount_Value	Total number of partitions in the cluster	count
kafka_controller_GlobalTopicCount_Value	Total number of topics in the cluster	count
kafka_partition_LogEndOffset	Last offset of data currently stored in the partition	count

Topic throughput

Metric name	Description	Unit
kafka_topic_BytesInPerSec_Count	Amount of data received by the topic per second	bytes/s(IEC)
kafka_topic_BytesOutPerSec_Count	Amount of data sent from the topic per second	bytes/s(IEC)
kafka_topic_MessagesInPerSec_Count	Number of messages received by the topic per second	count/s
kafka_topic_TotalFetchRequestsPerSec_Count	Number of Fetch requests per second	count/s
kafka_topic_TotalProduceRequestsPerSec_Count	Number of Produce requests per second	count/s
kafka_topics_FailedFetchRequestsPerSec_Count	Number of failed Fetch requests per second	count/s
kafka_topics_FailedProduceRequestsPerSec_Count	Number of failed Produce requests per second	count/s

Advanced Managed Search metrics

Primary metrics for monitoring nodes, indexes, search processing, and resource status of Advanced Managed Search (OpenSearch) clusters. These can be used in:

• Monitoring: Metric Export
• Alert Center: Setting metric-based notification policies

JVM and process

Metric name	Description	Unit
opensearch_process_cpu_percent	OpenSearch process CPU usage	%
opensearch_jvm_mem_heap_used_percent	Heap memory usage	%
opensearch_jvm_mem_heap_used_bytes	Amount of heap memory used	bytes(IEC)
opensearch_os_mem_used_bytes	Amount of OS memory used	bytes(IEC)
opensearch_jvm_gc_collection_count	Cumulative number of garbage collections	count
opensearch_circuitbreaker_tripped_count	Number of times the circuit breaker was triggered	count
opensearch_threadpool_tasks_number	Number of thread pool tasks	count
opensearch_threadpool_threads_count	Number of thread pool threads	count

Disk and storage

Metric name	Description	Unit
opensearch_node_disk_usage_percent	Node disk usage	%
opensearch_fs_total_total_bytes	Total disk capacity of all mount points	bytes(IEC)
opensearch_cluster_routing_allocation_disk_watermark_low_bytes	Low disk watermark capacity threshold	bytes(IEC)
opensearch_cluster_routing_allocation_disk_watermark_low_pct	Low disk watermark usage threshold	%
opensearch_cluster_routing_allocation_disk_watermark_high_bytes	High disk watermark capacity threshold	bytes(IEC)
opensearch_cluster_routing_allocation_disk_watermark_high_pct	High disk watermark usage threshold	%
opensearch_cluster_routing_allocation_disk_watermark_flood_stage_bytes	Flood-stage disk watermark capacity threshold	bytes(IEC)
opensearch_cluster_routing_allocation_disk_watermark_flood_stage_pct	Flood-stage disk watermark usage threshold	%

Cluster

Metric name	Description	Unit
opensearch_cluster_status	Cluster status (0: Green, 1: Yellow, 2: Red)	none
opensearch_cluster_nodes_number	Number of nodes	count
opensearch_cluster_datanodes_number	Number of data nodes	count
opensearch_cluster_shards_number	Number of shards by status	count
opensearch_cluster_shards_active_percent	Percentage of active shards	%

Index and search

Metric name	Description	Unit
opensearch_indices_doc_number	Current number of documents	count
opensearch_indices_indexing_index_count	Cumulative number of indexed documents	count
opensearch_indices_indexing_index_failed_count	Cumulative number of document indexing failures	count
opensearch_indices_search_query_count	Cumulative number of shard search query executions	count
opensearch_indices_search_query_time_seconds	Cumulative shard search query execution time	seconds

Direct Connect metrics

Primary metrics for monitoring traffic and connection status of Direct Connect virtual interfaces. These can be used in:

• Monitoring: Metric Export

Network traffic

Metric name	Description	Unit
dx_virtual_interface_input_bits_persec	Bits received per second on the virtual interface	bits/s(IEC)
dx_virtual_interface_output_bits_persec	Bits sent per second on the virtual interface	bits/s(IEC)
dx_virtual_interface_input_packets_persec	Packets received per second on the virtual interface	packets/s
dx_virtual_interface_output_packets_persec	Packets sent per second on the virtual interface	packets/s

Connection

Metric name	Description	Unit
dx_virtual_intrerface_state	Connection state of the Direct Connect virtual interface	count

Gateway Load Balancer metrics

Primary metrics for monitoring traffic, connections, and health status of Gateway Load Balancers and Endpoint Services. These can be used in:

• Monitoring: Metric Export

Network traffic

Metric name	Description	Unit
gwlb_bytes_in_persec	Total bytes received by the Gateway Load Balancer per second	bytes/s(IEC)
gwlb_bytes_out_persec	Total bytes sent by the Gateway Load Balancer per second	bytes/s(IEC)
eps_bytes_in_persec	Total bytes received by the Endpoint Service per second	bytes/s(IEC)
eps_bytes_out_persec	Total bytes sent by the Endpoint Service per second	bytes/s(IEC)
ep_bytes_in_persec	Total bytes received by the Endpoint per second	bytes/s(IEC)
ep_bytes_out_persec	Total bytes sent by the Endpoint per second	bytes/s(IEC)

Connections and endpoints

Metric name	Description	Unit
gwlb_current_connections	Number of active connections on the Gateway Load Balancer	count
gwlb_healthy_host_count	Number of healthy hosts on the Gateway Load Balancer	count
gwlb_unhealthy_host_count	Number of unhealthy hosts on the Gateway Load Balancer	count
eps_current_connections	Number of active connections on the Endpoint Service	count
eps_endpoint_count	Number of endpoints connected to the Endpoint Service	count
ep_current_connections	Number of active connections on the Endpoint	count

Private Endpoint metrics

Primary metrics for monitoring traffic and connection status of Private Endpoints. These can be used in:

• Monitoring: Metric Export

Network traffic

Metric name	Description	Unit
ep_bytes_in_persec	Total bytes received by the Endpoint per second	bytes/s(IEC)
ep_bytes_out_persec	Total bytes sent by the Endpoint per second	bytes/s(IEC)

Connections

Metric name	Description	Unit
ep_current_connections	Number of active connections on the Endpoint	count