Understanding the memory usage in MMAPv1 is relatively straightforward. MMAPv1 uses the mmap() system call under the hood to pass on the responsibility of the memory page to the underlying operating system. That is why when we use MMAPv1, the memory usage will grow unbounded, as the operating system is trying to fit as much of our dataset into the memory as possible.

With WiredTiger, on the other hand, we define the internal cache memory usage on startup. By default, the internal cache will be, at a maximum, between half of our RAM, which is in-between 1 GB or 256 MB.

On top of the internal cache, there is also memory that MongoDB can allocate for other operations, like maintaining connections and data processing (in-memory sort, MapReduce, aggregation, and more).

MongoDB processes will also use the underlying operating system's filesystem cache, just like in MMAPv1. The data in the filesystem cache is compressed.

We can view the settings for the WiredTiger cache via the mongo shell, as follows:

> db.serverStatus().wiredTiger.cache

We can adjust its size by using the storage.wiredTiger.engineConfig.cacheSizeGB parameter.

The generic recommendation is to leave the WiredTiger internal cache size at its default. If our data has a high compression ratio, it may be worth reducing the internal cache size by 10% to 20% to free up more memory for the filesystem cache.