[MUD-Dev] Disk v. Mem

[Miroslav Silovic]

[Charset ISO-8859-1 unsupported, filtering to ASCII...]
> 
> erh, at the risk of getting bounced out of
> here for asking a fourth-grade question... ;) 
> 
> what are the advantages of swapping out objects
> to disk as opposed to keeping them in memory?
> 
> let me stab in the dark:
> do you really need to save that much memory? 
> do you need it for something else? 
> are objects so large that 40,000 of them fill mem quickly?
> does too full a mem size inhibit the driver speed?
> isn't it more costly in speed to write to disk? 
> what kind of savings do you get versus this extra cost?

The biggest Cold database constructed so far had about 150,000 objects.
That was 200MB *ON DISK*.

Trying to load all of this in RAM would trash your swap, blast away your
paging system, burn your village, rape your women, kill your men and
children, .... :}

Incidentally, the biggest MUSH database I've seen was about 120MB (old
PernMUSH). About 90% of this was just plain text.

And no, it is not more costly to write it to disk. Remember, your OS
doesn't know the object boundaries, so it just swaps off the entire
pages. And one object may touch a *lot* of pages. So, once object is
paged out, OS may need to load, say, 4-5 pages to get to a single
object. In most UNIces this means 20 killos.

> i'm seriously inquiring here...before i've just let the
> whole subject go by without thinking on it too much.
> 
> -john
> 
>