We are currently using MySQL RDS (v8) and have recently adjusted our maintenance window from 8AM UTC on Sundays to 4AM UTC. A Version Upgrade was also available for our instance. Via the RDS interface, we opted for this update to be applied during our next maintenance window. Despite us selecting the "Apply on next maintenance window" option, the system proceeded to update mode, and after about 30 minutes, it appeared that our RDS instance had been updated. This event took place today, on Sunday, May 21, 2023, at 10PM UTC, which is outside of our designated maintenance window. We were cautious not to click the "Apply immediately" option to avoid any downtime or availability issues during non-maintenance hours.
However, this unexpected upgrade resulted in a 15-minute downtime for our application. We are puzzled as to why this upgrade was executed immediately, rather than waiting for the next scheduled maintenance window (Sunday, May 28, 2023, at 4PM UTC).
This unexpected behaviour of the system has led to some disruptions in our application, and we seek an explanation for this occurrence.
AWS OFICIALAtualizada há 6 meses