Yeah, you did misunderstand. The name of this feature is unfortunately not very self-explanatory.
Basically, it’s the opposite of your interpretation. You don’t put cache/RAM contents onto the disk, but rather keep disk contents in cache/RAM.
When an application reads from disk, this portion of the disk – a so-called “page” – gets loaded into RAM. When the application is done with this page, the OS doesn’t overwrite it with zeros, but rather just keeps it in there until the memory is actually needed for something else.
And if that same page is needed again, the OS can skip the hard drive access, which massively speeds up access.
Disk ram doesn’t sound like a good idea for modern limited SSD read/writes. -Or do I misunderstand?
Yeah, you did misunderstand. The name of this feature is unfortunately not very self-explanatory.
Basically, it’s the opposite of your interpretation. You don’t put cache/RAM contents onto the disk, but rather keep disk contents in cache/RAM.
When an application reads from disk, this portion of the disk – a so-called “page” – gets loaded into RAM. When the application is done with this page, the OS doesn’t overwrite it with zeros, but rather just keeps it in there until the memory is actually needed for something else.
And if that same page is needed again, the OS can skip the hard drive access, which massively speeds up access.
Here’s the Wikipedia article: https://en.wikipedia.org/wiki/Page_cache
Seems a common issue in Linux, like with display manager / stable. lol -thanks for clarifying!
I had sent my reply too early. You’ll want to read the edit. 🫠
I could tell, I do the same. Excellent reply. My head is spinning from reviewing all this old stuff in a marathon way, and you made it clear! -Thanks!