>DP> hmm... what i tried is a loop device of a fixed size (100mb) on the 
>DP> directory that contains small files. it works fine, but is a little 
>DP> bit of a pain changing size. 

200mb will be enough)

/var/lib/pacman-cage.img  194M   40M  145M  22% /var/lib/pacman

>DP> conceptually seen, small files should not be stored plainly on the fs. 
>DP> one idea would be to have the fs to decide what is small file what a 
>DP> big file and what has the potential to fragment and then on the back 
>DP> reshuffle things and change strategy of storage to optimise access. 
>DP> it would be like a library or an archive (with books in it - the real 
>DP> library kind of library *g*) where the people working there sort 
>DP> things by its properties. if single papers are filed, they end in 
>DP> folders. if encyclopedias are filed, whole shelfs are reserved for 
>DP> them. if a paper develops in an encyclopedia, then the person working 
>DP> there changes things. the CPU's nowadays are capable of providing the 
>DP> power (in the background) to do such intelligence (complex 
>DP> algorithms) but i am not aware of an implementation on the fs-level. 
>DP> there are some sql-like database-like approaches (winfs?) but this is 
>DP> still not facing the problem directly. 

>DP> we need to employ somebody to tell the fs driver where to store 
>DP> things... anybody knows a librarian with free time? 'Oook' :)

As I know xfs have some db api...

But about pacman - may be db engine should be added into it? For example
own engine(not berkley db nor sqlite)?

_______________________________________________
arch mailing list
[email protected]
http://archlinux.org/mailman/listinfo/arch

Reply via email to