Fault-tolerant principle

Hardware, software, and human fault tolerance should be part of this pattern. The pattern is for catering to big data, and because of this, any of these faults can be a big problem to recover from. So, data loss and data corruption don’t have any  place in this pattern because of the data vastness. If it does have this, in most cases, it is irrecoverable; so this principle is quite a strong need.

One of the important parts of this is human fault tolerance. Some of the very common mistakes are typical operational mistakes made in day-to-day operations; the next most common mistakes are bugs introduced over a period of time (releases) into the production system. The system should indeed be designed to cater to and address these aspects.