Some background - we're going to move to a Redis cluster for caching and we obviously want to test that...
- It works at all
- What happens if an individual node is removed
- Is the performance better/worse/the same.
...the trouble is that I'm struggling to come up with a good way to achieve this. I don't think "we'll try a few pages and see what happens" will be acceptable, but variations on that are really all I have.
Does anybody have any thoughts or, even better, has done this?