Two days ago our MySQL RDS instance saw a sudden spike in BinLogDiskUsage, CPUUtilization, NetworkReceiveThroughput and NetworkTransmitThroughput.
The result was a database that was extremely slow to respond, but not unresponsive. Requests to the APIs connected to the database were taking 10x longer to respond than normal.
The WAF did not indicate an increase in traffic (although it did coincide with what is typically the busiest part of the day).
Within three hours, the situation resolved itself, and response times returned to normal.
My questions are:
- What would cause BinLogDiskUsage to suddenly increase?
- Why would this cause the DB to become so slow, as CPUUtilization only reached around 20%? (Am I confusing cause and effect?)
If it's helpful, the