All Products
Search
Document Center

ApsaraDB for ClickHouse:Metric descriptions

Last Updated:Mar 28, 2026

ApsaraDB for ClickHouse provides metrics that let you monitor cluster health in real time, detect issues early, and locate performance bottlenecks quickly.

The metrics in the Community Edition and Enterprise Edition sections apply to clusters purchased after December 1, 2021. For clusters purchased before December 1, 2021, see Metrics for earlier cluster versions.

Community Edition

MetricUnitDescription
Disk throughputByteRead and write throughput (bandwidth) of the disk.
Disk IOPScounts/sNumber of read and write operations performed on the disk per second.
Disk usageByteAmount of disk space used.
Disk usage - by single nodeByteTotal disk space used on a single node.
Disk usage%Ratio of used disk space to total available disk space.
Disk usage - by single node%Disk usage calculated per node.
CPU utilization%Average CPU utilization of each node in the cluster.
inode usage%Ratio of used inodes to total available inodes. An inode is a Linux data structure that identifies a file or directory.
Memory usageByteMemory used by each node in the cluster.
Memory utilization%Average memory utilization of each node in the cluster.
Number of data partscountTotal number of data parts across the cluster.
Number of inactive data partscountNumber of data parts that are no longer active.
ZooKeeper CPU utilization%Average CPU utilization of the ZooKeeper nodes in the cluster.
ZooKeeper memory utilization%Average memory utilization of the ZooKeeper nodes in the cluster.
Write bytes/secByteVolume of data written to each node per second.
Write rows/secrows/sNumber of rows written to each node per second.
QPScountNumber of queries processed per second.
TPScountNumber of transactions processed per second.
Network throughputByteNetwork read and write throughput (bandwidth).
TCP connectionscountNumber of TCP connections to the cluster.
HTTP connectionscountNumber of HTTP connections to the cluster.
Number of running Init QueriescountNumber of queries running locally on the node, not forwarded from another node.
Number of running queriescountTotal number of queries currently running on the node.
Number of failed queriescountNumber of queries that failed.
Number of failed Insert queriescountNumber of INSERT statements that failed.
Number of delayed insertscountNumber of INSERT statements that were delayed.
Number of failed Select queriescountNumber of SELECT queries that failed.
Number of running mergescountNumber of background merge tasks currently running.
Number of running mutationscountNumber of data correction tasks (mutations) currently running.
Number of temporary files for distributed tablescountNumber of temporary files created for distributed tables.
Number of MaterializeMySQLcountNumber of databases created using the MaterializeMySQL engine.
Cold storage usageByteAmount of cold data currently stored.
Number of failed MaterializeMySQL synchronizationscountNumber of failed sync operations for databases created with the MaterializeMySQL engine.
Number of errors in Kafka external table consumptioncountNumber of consumption failures for Kafka external tables.
ZooKeeper Leader-Indicates the ZooKeeper leader node. A value of 1 identifies the leader.
Number of Kafka external tablescountNumber of Kafka external tables created in the cluster.
Node queue accumulationcountNumber of data packets in the send queue of an ApsaraDB for ClickHouse node on ZooKeeper.
ZooKeeper average latencymsAverage request latency of the ZooKeeper nodes.
Number of ZooKeeper nodescountNumber of ZNodes in ZooKeeper.
ZooKeeper TPScounts/sTransactions per second processed by ZooKeeper.
Zxid usage%Zxid usage of ZooKeeper. When Zxid usage reaches 100%, a new leader election is triggered.

Enterprise Edition

MetricUnitDescription
CPU utilization%Average CPU utilization of each node in the cluster.
CCUcountAmount of compute resources (CCU) used by the cluster.
Memory usageByteMemory used by each node in the cluster.
Memory utilization%Average memory utilization of each node in the cluster.
OSS usageByteAmount of OSS storage used by the cluster.
Number of data partscountTotal number of data parts across the cluster.
Number of inactive data partscountNumber of data parts that are no longer active.
Keeper CPU utilization%CPU utilization of ClickHouse-Keeper.
Keeper memory usageByteMemory used by ClickHouse-Keeper.
Keeper memory utilization%Memory utilization of ClickHouse-Keeper.
Write bytes/secByteVolume of data written to each node per second.
Write rows/secrows/sNumber of rows written to each node per second.
Network throughputByteNetwork read and write throughput (bandwidth).
OSS network throughputByteRead and write throughput (bandwidth) of OSS.
OSS IOPScounts/sNumber of read and write operations performed on OSS per second.
TPScounts/sNumber of transactions processed per second.
QPScounts/sNumber of queries processed per second.
Number of running Init QueriescountNumber of queries running locally on the node, not forwarded from another node.
Number of running queriescountTotal number of queries currently running on the node.
Number of failed queriescountNumber of queries that failed on the node.
Number of failed Insert queriescountNumber of INSERT statements that failed.
Number of delayed insertscountNumber of INSERT statements that were delayed.
Number of failed Select queriescountNumber of SELECT queries that failed on the node.
Number of running mutationscountNumber of data correction tasks (mutations) currently running.
Number of running mergescountNumber of background merge tasks currently running.
TCP connectionscountNumber of TCP connections to the cluster.
HTTP connectionscountNumber of HTTP connections to the cluster.
Number of failed MV refreshescountNumber of failed refresh operations for refreshable materialized views in the cluster.
Average MV refresh durationseconds/minutesAverage time to complete a refresh of refreshable materialized views in the cluster.
Maximum MV refresh durationseconds/minutesMaximum time to refresh refreshable materialized views in the cluster.

Metrics for earlier cluster versions

Earlier cluster versions refer to Community Edition clusters purchased before December 1, 2021.
MetricUnitDescription
CPU utilization%CPU utilization of each node in the cluster.
Memory utilization%Memory utilization of each node in the cluster.
Memory usageByteMemory used by each node in the cluster.
Disk usage ratio%Ratio of used disk space to total available disk space.
Disk usageByteAmount of disk space in use.
Disk IOPScounts/sNumber of read and write operations performed on the disk per second.
Disk IOPS sizeByteAmount of data read from or written to the disk per second.
Database connection usage ratio%Percentage of used database connections out of the maximum allowed.
Used database connectionscountNumber of database connections currently in use.
TPScounts/sNumber of transactions processed per second.
Write rows/secrows/sNumber of rows written to each node per second.
Write bytes/secByteAmount of data written to each node per second.
QPScounts/sNumber of queries processed per second.
Average ZooKeeper wait timemsAverage response latency of ZooKeeper.
Average I/O wait timemsAverage I/O response latency.
Average CPU wait timemsAverage CPU response latency.