Re: [PATCHSET v3 0/2] mkfs: stop allowing tiny filesystems

["Darrick J. Wong" <djwong@kernel.org>]

On Mon, Jul 25, 2022 at 09:59:42AM +0200, Carlos Maiolino wrote:
> On Tue, Jul 19, 2022 at 02:44:58PM -0700, Darrick J. Wong wrote:
> > Hi all,
> > 
> > The maintainers have been besieged by a /lot/ of complaints recently
> > from people who format tiny filesystems and growfs them into huge ones,
> > and others who format small filesystems.  We don't really want people to
> > have filesystems with no backup superblocks, and there are myriad
> > performance problems on modern-day filesystems when the log gets too
> > small.
> > 
> > Empirical evidence shows that increasing the minimum log size to 64MB
> > eliminates most of the stalling problems and other unwanted behaviors,
> > so this series makes that change and then disables creation of small
> > filesystems, which are defined as single-AGs fses, fses with a log size
> > smaller than 64MB, and fses smaller than 300MB.
> > 
> > v2: rebase to 5.19
> > v3: disable automatic detection of raid stripes when the device is less
> >     than 1G to avoid formatting failures
> > 
> > If you're going to start using this mess, you probably ought to just
> > pull from my git trees, which are linked below.
> > 
> > This is an extraordinary way to destroy everything.  Enjoy!
> > Comments and questions are, as always, welcome.
> > 
> 
> Both changes looks good to me, but IMHO we really require it to be documented in
> manpages otherwise we'll get (even more) questions about "why can't I create
> small FS'es anymore?".
> But anyway, I can help with the manpages once these patches hit for-next if you
> are ok with it.

Oooh, good point, Eric and I have been too busy figuring out the weird
corner cases and forgot that.  I'll add some manpage updates and send
that out tomorrow.

--D

> 
> -- 
> Carlos Maiolino