Fine. If the pg team want to call their shared memory space a disk
buffer, let them. And you can too. Anything committed to disk still
has to traverse the os disk cache. So in reality, it depends upon how
you balance parameters such as your os disk cache and your sql disk
cache etc etc. I think we've now flogged this enough now.
This is absolute nonsense. The shared buffer cache is well understood,
and the only way you will hurt performance by making it too big is by
using up so much RAM that you start hitting swap, or by making it
larger than your data plus the other usage. Unless perhaps your OS
performs poorly with large shared memory allocations (does openbsd?).
Which is what the original poster asked, because "they" are saying that
FreeBSD's shared memory management is superior (compared with what?)
So, does OpenBSD page shared memory to disk if ram becomes full? If it
does, can it be prevented?
G
