Madelyn Olson

https://github.com/redis/redis/actions/runs/5301569607

I think this should fix it: https://github.com/redis/redis/pull/12330.

I have no concerns, the actual change to the core seems small.

@oranagra the only item left here is "Redis does not fail if a category or command is missing", do we still think that is needed? I think the current acl...

> The most significant risk I see is an incorrect rule addressing a non-existing command/group (thereby an NOP) that becomes effective due to group/command changes. But it's too theoretical and...

> I assumed it's fully resolved, not mostly resolved. If that's not the case, we should avoid it. Yeah, I think so. I'm just very hesitant to change the expectations...

@avacd Redis is not smart enough to tell that, it doesn't introspect into the pattern to see that it contains a hash tag. It seems pretty straight forward to add...

@avacd I'm going to propose an alternative for you that I've been thinking about. I want to logically store keys with the same tag together and allow doing flushtag {tag}...

Well, I was very optimistic about the "almost ready" part. It's a lot harder than we thought :/, but we'll post it when we can. One tenet I would like...

So we have three use cases: * I need a consistent view of all keys in a single slot. The main use case is for slot migration. `cluster getkeysinslot` covers...