MySQL Monitoring

Graphs

MySQL clusters provide information about the key performance indicators of the cluster nodes – vCPU, memory, and disk load.

vCPU

The graph shows the percentage of the cluster server cores loaded.

Load Average

The graph shows the average system load over a period of time. Load average shows the number of processes which are being executed by the cluster cores. The graph shows the indicator displayed as three values: for 1 minute, for 5 minutes, and for 15 minutes.

Memory Usage

The graph shows the taken memory without regard of the cache and OS buffers.

Disk Usage

The graph shows the used disk space.

Disk Fullness Notifications

Overview

When the disk is 80% full, the service automatically sends a notification to the email associated with the account that the disk is almost full. If it is assumed that the load on the cluster will not decrease or will continue to grow, you should scale the cluster (choose a configuration with a larger disk size). Alternatively, you can clean the disk by running autovacuum or deleting unnecessary databases. If the cluster disk runs out of space, the cluster’s status will change to Disk_full. In this state, the cluster works only for read – no write operations are performed.

Please note that the cluster’s status changes to Disk_full when the disk is 95% full.

For the cluster to start working again for write operations, you need to:

  • clean the cluster disk – you can do it yourself or contact our technical support service — we will clean the disk;
  • scale the cluster – select a configuration with a larger disk size.

Please note that for the cluster to be able to perform read and write operations again, the disk must be 85% or less full.

Learn more about the history of operations and cluster statuses in the Managing clusters and databases article.