Introduction to Monitoring Metrics

Monitoring Metrics

Description

CPU

CPU usage of an instance, including: sys: CPU usage in kernel mode, usr: CPU usage in user mode.

Memory

Memory utilization of an instance, including: Machine memory utilization, specific memory usage of database.

Disk

Disk performance metrics related to an instance, including disk utilization, Iops: input/output operations per second, read/write traffic per second (kb), and average disk queue length.

Connection

Number of connections of an instance.

TPS-QPS

Read/write performance metrics of the database, including:

command operands, query operands, insert operands, delete operands, update operands, and getmore operands.

Network

Network traffic of an instance.

Global lock

Length of the queue waiting for global locks, including the length of the queue waiting for global read lock, the length of the queue waiting for global write lock, and the length of the queue waiting for all global locks.

Cursor

The number of cursors currently in use by an instance, including the number of open cursors and timeout cursors.

Number of page faults

Rate of processing the fault pages per second.

Assertion

Including: average increasing regular assertion rate per second, average warning rate per second, average message rate per second, and average user assertion rate per second.

wiredtiger

Storage engine related metrics, including:

wiredtiger memory data volume, wiredtiger memory dirty data volume, wiredtiger frequency of writing data into memory, and wiredtiger frequency of writing data from memory into disk.

Oplog (unique to replica sets)

Including available oplog time of the master node, oplog generation rate of the master node, active/standby oplog overlapping duration, and active/standby delay.