On Sat, 22 Aug 1998 17:52:34 -0500, Jeff Noxon <jeff@planetfall.com>
said:
> On Sat, Aug 22, 1998 at 06:56:22PM +0100, Stephen C. Tweedie wrote:
>> We do have prototype code to allow multiple devices to be bound together
>> at the filesystem layer. That will allow us to go above 1TB per
>> filesystem, but we'll still be limited to 8TB because of the maximum fs
>> blocksize of 4096 bytes. However, Ted's btree index code, when ready,
>> will allow us to get past even that limit. :)
> Why is the block size limited to 4KB? It seemed to me that 8KB might
> be a better choice on the Alpha for performance-oriented stuff, since
> the MM page size is 8KB. Is this an e2fs limit or a kernel limit?
It's an ext2 limit. ext2fs is careful to use a portable on-disk
format, swapping byte orders when necessary for example. An ext2fs
zip cartridge formatted on any Linux system should be usable on any
other. One consequence is that we limit the block size to the lowest
common denominator of page sizes, which is 4096 bytes. An 8192-byte
blocksize filesystem would not be portable.
--Stephen
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html