Supported Metrics

Description

This section describes metrics reported by the Data Replication Service (DRS) to Cloud Eye as well as their namespaces and dimensions. You can use APIs provided by Cloud Eye to query the metrics of the monitored object and alarms generated for DRS.

Namespace

SYS.DRS

DB Instance Monitoring Metrics

Table 1 lists the DRS performance metrics.

**Table 1** DRS metrics
Metric ID	Metrics Name	Description	Value Range	Monitored Object	Monitoring Interval (Raw Data)
cpu_util	CPU Usage	CPU usage of the monitored object	0-100%	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
mem_util	Memory Usage	Memory usage of the monitored object	0-100%	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
network_incoming_bytes_rate	Network Input Throughput	Incoming traffic in bytes per second	≥ 0 byte/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
network_outgoing_bytes_rate	Network Output Throughput	Outgoing traffic in bytes per second	≥ 0 byte/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
disk_read_bytes_rate	Disk Read Throughput	Number of bytes read from the disk per second (bytes/second).	≥ 0 byte/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
disk_write_bytes_rate	Disk Write Throughput	Number of bytes written to the disk per second (bytes/second).	≥ 0 byte/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
disk_util	Storage Space Usage	Storage space usage of the monitored object	0-100%	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
extract_bytes_rate	Source Database Read Throughput	Table data or WAL bytes read from the source database per second	≥ 0 byte/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
extract_rows_rate	Rows Read from Source Database per Second	Number of table data rows or WAL rows read from the source database per second Unit: rows/s.	≥ 0 row/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
extract_latency	Source Database WAL Extract Lag	Latency of extracting WAL from the source database Unit: ms.	≥ms	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_bytes_rate	Destination Database Write Throughput	Number of bytes written to the destination database per second.	≥ 0 byte/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_rows_rate	Rows Written into Destination Database per Second	Number of rows that are written to the destination database per second Unit: rows/s.	≥ 0 row/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_transactions_rate	DML TPS	Number of DML transactions written to the destination database per second.	≥ 0 transaction/s	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_ddls_numberor apply_ddls_rate NOTE: apply_ddls_rate is replaced by apply_ddls_number after December 2022.	DDL TPS	Total number of DDL transactions written into the destination database.	≥ 0 transaction	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_latency	Replication Delay	Delay (in milliseconds) of data replay.	≥ 0 ms	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_average_execute_time	Average Transaction Execution Time	Average execution time (RT = Execution time + Commit time) of a transaction in the destination database. The unit is millisecond.	≥ 0 ms	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_average_commit_time	Average Transaction Commit Time	Average commit time (RT = Execution time + Commit time) of a transaction in the destination database. The unit is ms.	≥ 0 ms	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_current_state	Synchronization Status	This metric is the synchronization status of the current kernel data (10: abnormal; 1: idle; 2: DML; 3: DDL), instead of the task status.	10: abnormal 1: idle 2: DML is executed. 3: DDL is executed.	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_thread_workers	Synchronization Threads	Number of working threads for data synchronization	≥ 0	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute
apply_job_status	Task Status	Status of the current task. (0: normal; 1: abnormal; 2: paused)	0: normal 1: abnormal 2: paused	Monitored object: ECS Monitored instance type: replication, synchronization, and DR instances	1 minute