> but its incredibly costly and time consuming for sizeable platforms.
If your restores are too time consuming to test regularly, they sure as shit aren't going to be useful in a disaster.
Why is that? Hours or even days of downtime are still better than just losing all data. It's a simple cost-benefit analysis, and it's ok to pick different trade-offs depending on your use case
It depends on the accepted downtime.
Why is that? Hours or even days of downtime are still better than just losing all data. It's a simple cost-benefit analysis, and it's ok to pick different trade-offs depending on your use case