You need the platform to handle ten times today's customers. The scheduling engine can. This database can't, so that's only true if this work is done. The business plan reaches three times in month 14.
Evidence
Three of the nine pieces of evidence. Each was checked again against the code before release.
- Resource
- rds / kestrel-prod
- Read-only copies
- None
- Standby in a second zone
- Yes
- Used by
- scheduler, api, reporting
Management's reply
We accept the finding. A read replica is planned for the first quarter. We dispute the threshold: our own estimate is that this bites at five times, not three.
The planned replica is noted. If it's delivered before completion, the likely cost falls by £45k. The threshold stays at three times: month-end load is already 86 to 91%, and no evidence was given for five.
Ask a question
About £45k of the £180k. It fixes the month-end slowdown but not the limit on growth. That needs reporting moved to its own database.
The evidence in this report doesn't answer that. Your question went to the analyst on 8 Nov at 16:40. The reply will appear here.
With the analyst