Partition encompasses response time. If you have defined SLAs on response time and your partition times exceed that, you're not available.

One option to increase availability though is to reduce partition times.

> The CAP theorem is also irrelevant if your acceptable response time is greater than the time it takes your partitions to sync. This is really an oversimplification. The more important metric here is the delay between write and read of the same data. Even in that case if when the system write load is unpredictable it will definitely lead to high variance in replication lag. The number of times I had to deal with a race condition for not considering replication lag factor is more than I would like to admit.

What's the difference between choosing an acceptable response time that is greater than the time it takes for your partitions to sync and giving up on availability?

I don't think it makes sense to say that CAP doesn't apply if you don't need consistency, availability, or tolerance to partitions. CAP is entirely about the need to relax at least one of those three to shore up the others.

You mean CLAP? Latency and availability are basically the same thing. CAP is simple.

No I'm talking about response time 'requirements'

and considering most cloud compute has a better backbone or internal routing, userspace should see this less.

That being said, if this is truly a problem for you CRDB is basically built with this all in mind.

Wow have heard about CRDB but just now reading what it does.

That's an incredible piece of engineering.

I mean you just KNOW that while that addition may make sense, some 12 year minded person like me would just start referring to it as the CRAP theorem. And I don’t even dislike the theorem.