Ted T'so on SSD and journalling

[Alan Dayley]

2009/6/23 der.hans <PLUGd at lufthans.com>
>
> moin moin,
>
> The first paragraph in the quote was a comment on Ted's blog. The second
> is Ted's response. Read the blog post at the URL given below.
>
> ###
> #  # 2  tytso  Says:
> March 2nd, 2009 at 1:15 am
>
> @1: I’d thought that the reason to avoid ext3 on SSDs, at least most of
> the ones available today, was not the total number of writes but rather
> the repeated writes to the same place on the disk (that is, the journal)
>
> Norman,
>
> Actually, even the most primitive SSD’s and Flash drives have to get this
> right, because the Windows FAT filesystem are constantly updating the same
> locations on disk (namely for the File Allocation Table), which is in a
> fixed location on disk. So although there’s not a lot we can count on in
> terms of the quality of flash drives’ wear level, it’s very likely they
> get that right, since otherwise their reliability on basic FAT
> filesystems, which are used in essentially every single digital camera on
> the market, would be pretty bad.
> ###
>
> http://thunk.org/tytso/blog/2009/03/01/ssds-journaling-and-noatimerelatime/

Interesting post!

Several years back in a discussion with a flash controller maker we
discussed the FAT area of their system.  This particular controller
would set aside a area of the LBAs for use of the tables for a FAT
file system.  The flash erase blocks were smaller, wear-leveled more
often, etc.

This is good for a system that uses a FAT file system.  It is not good
for many other file systems that have high write areas in other than
the low LBAs.  The controller maker was surprised when when we pointed
this out to them, noting that there are a great many Unix-based file
systems in use for the embedded, industrial and military markets.

Many controllers now days do maximum wear-leveling across the entire
flash space so the distinction of where a FAT is located is less
important.  This is good news for ext3 and other journaling file
systems.  Running ext3 on a modern SSD should not cause more wear-out
than what is proportional to the total number of writes experienced.

Alan