Re: Bug#605009: serious performance regression with ext4

[Olaf van der Spek <olafvdspek@gmail.com>]

On Fri, Nov 26, 2010 at 3:53 PM, Raphael Hertzog <hertzog@debian.org> wrote:
> It all started with report of corrupted (zero-length) files on ext4/ubifs
> (see http://bugs.debian.org/430958). We did the right thing to fix this
> which is to call fsync() on the fly on each file that dpkg unpacks. That

Is durable required or is atomic enough? The fact that fsync wasn't
present suggests durable isn't required.

IMO the ext4 issue should be solved by supporting atomic file updates
without all kludges like renaming.

Olaf