Re: [tlug] SSD filesystem

[Clemens Schwaighofer <gullevek@example.com>]

On Fri, Jun 5, 2009 at 21:21, Francois Cartegnie<fcartegnie@example.com> wrote:
> Le mercredi 3 juin 2009, Kyle Hasegawa a écrit :
>> Seems everyone has an opinion on which filesystem and settings are best
>> for an SSD drive.  Some say journaling filesystems like ext3 will wear
>> out the drive sooner though opponents argue that failure from
>> overwriting is exaggerated.  Others recommend ext4 to get even better
>> performance out of a high speed SSD.
> Hi,
>
> That's still a discussion for getting maximum performance and reliability from
> a device where the blocks wear levelling is mainly optimized for FAT or NTFS
> filesystems.
>
> Some people are waiting for devices allowing raw access to flash memory, able
> to shortcut the wear levelling algorithms. Emerging SDD filesystems are still
> relying on an "unknown strategy" for wear levelling, and have no way to
> implement their own :/

I wouldn't do that, I think the maker knows better where and how to
avoid bad blocks in SSD devices.

> But maybe someone here has heard of new SSD hardware, *nix fs friendly.

I guess they are all the same like HDs, who in the world wants to
override the automatic badblock system in a normal HD?

> I guess your ext4/ext3results will differ depending on the hardware used (also
> SLC/MLC), and comparison is even more difficult with the RAM embedded ones.

FS need to change for that, so there will be some major rewrite to
address the difference between oldschool HDs and SSDs, but that will
take some time. SSDs are by fair a common good. Pricey for the good
ones, and real crappy slow for the bad ones.

-- 
Clemens Schwaighofer
gullevek@example.com / gullevek@example.com
http://www.flickr.com/photos/gullevek/