Following a recent deployment, users experienced application slowdown and request failures. The slowdown was traced back to an internal change that affects how SQL queries are composed. To mitigate this incident, both a reversion of the internal change and an upscale of the database were executed.