[neovintage]

“Data Downtime” is very real and can have terrible impacts on understanding what’s happening to customers or your business. I like that this series is taking the simple approach of using SQL to help identify the problems. (Its also a good lesson on window functions) Run these queries on some kind of schedule and you’ve got a decent solution to monitor the situation. Granted there’s different levels of sophistication to monitor these problems with better tools as your data volumes grow. The bigger problem beyond downtime is how to recover from these situations. I’ve seen many databases architected in ways where you can’t recover said downtime. That’s where change data capture could really help but I think that’s beyond the scope of this series.