Alex Rousskov wrote:
> I would be cautious about this. If many requests for different objects
> go to the same disk, then it does not matter much whether subsequent
> requests for parts of the same object go to the same place on disk.
Assuming each block of data is large enought to allow efficient use of
disk/fs prefetching.
/Henrik
Received on Wed Dec 13 2000 - 22:06:37 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:13:04 MST