RAM is fast, yet expensive. Hard disk drives, or solid state drives, are relatively cheaper and slower, and also provide durability for our data in the case of system and power failures. All of our data is stored on the disk, and when we perform a query, MongoDB will try to fetch data from memory. If the data is not in the memory, then it will fetch the data from the disk and copy it to the memory. This is a page fault event, because the data in the memory is organized in pages.

As page faults happen, the memory gets filled up, and eventually, some pages need to be cleared for more recent data to come into the memory. This is called a page eviction event. We cannot completely avoid page faults unless we have a really static dataset, but we do want to try to minimize page faults. This can be achieved by holding our working set in memory.