How does FoundationDB's distributed caching works?

I found that FoundationDB already has its own distributed caching implementation from here.
This introduction is too simple.
I hope to learn about some details information.
For example, what data is cached, where and when the cached data is stored, and how it ensure ACID properties.
Can anyone tell me where the related document is or just introduce the high level idea for me.
Thanks very much!
(By the way, is this feature enabled by default?

We don’t have a distributed cache feature (yet) and imho the documentation is slightly misleading. I think what this is referring to is the page cache: storage servers use the memory on their machine to cache recently read data. Depending on the workload and the amount of memory given to a process, this can be very efficient and make reads almost as fast as if you read from an in-memory database. However, it will still be slightly slower (reading from an on-disk data structure is typically slower than reading from an in-memory data structure, even if no disk reads are required).

1 Like