Scenarios

CPU spike

• The CPU usage remains close to or reaches 100%.
• The query response time increases significantly.
• The overall system performance deteriorates, which may cause stuttering or lag.

• As service traffic increases, a large number of concurrent query requests flood into the database.
• The number of complex SQL queries increases.
• The application-layer cache becomes invalid, causing a large number of requests to directly access the database.
• During peak hours, user traffic surges.
• The database index design is improper.

• The application response delay increases, and user experience deteriorates.
• The transaction processing capability deteriorates, and the service throughput decreases.
• The system stability deteriorates, and timeout errors may occur.
• In extreme cases, the service may be unavailable.

High memory load

• The memory usage remains high.
• The database cache hit ratio decreases.
• Frequent memory swapping may occur in the system.
• A memory usage alarm is generated.
• A primary/standby switchover occurs or the node reboots.

• As traffic increases, the database needs to cache more data and query results.
• The number of concurrent connections increases, and each connection occupies memory.
• A large number of complex queries increase the usage of session-level memory, such as the sort buffer (sort_buffer) and join buffer (join_buffer).
• Stored procedures generate a large amount of cached data for subsequent use, avoiding repeated parsing overhead. If a large number of different or complex stored procedures are called, cached data will quickly accumulate, which may cause a sharp increase in memory usage and affect the overall database performance.
• Accumulation of prepared statements causes the memory usage to increase slowly.

• The database performance deteriorates, and the query response time increases.
• An out of memory (OOM) error is triggered, causing a primary/standby switchover or node reboot. During the switchover or reboot, services are intermittently interrupted.

Connection exhaustion

• The database connection pool usage is close to or reaches the upper limit.
• The waiting time for new connection requests increases.
• Applications may experience database connection timeout errors.
• An alarm about connection exhaustion is generated.

• As service traffic increases, applications require more database connections.
• The connection pool configuration at the application layer is improper.
• A large number of requests flood in within a short period of time, causing the number of connections to increase rapidly.
• Increased traffic leads to longer transaction processing time and prolonged connection occupation.
• Connections are not correctly released due to application code issues.

• Applications cannot obtain database connections, resulting in request failures.
• The system throughput decreases and cannot process more concurrent requests.
• User experience deteriorates, and error messages may be displayed.
• In extreme cases, application services may be unavailable.