Quiescent filesystems marked with EXT2_VALID_FS while still mounted?

[Rogier Wolff]

tytso en MIT.EDU wrote:
>    From: David Woodhouse <dwmw2 en infradead.org>
>    Date: Wed, 19 Jan 2000 13:28:55 +0000
> 
>    I'm fairly sure that under 2.0 kernels, if a filesystem had been
>    unused for a long period of time, it would be marked with
>    EXT2_VALID_FS - with the highly desirable effect that if the machine
>    subsequently crashed, the filesystem wouldn't need a fsck.
> 
> Nope, it never did that.
> 
> You could potentially do something like that, but it would mean that
> whenever you tried to write to the filesystem (including inode atime
> updates), all modifications would have to stall until the EXT2_VALID_FS
> bit was cleared.  It's doable, but it's not clear it's worth it.

Yes, yes! It's worth it! That one extra "write" will take a seek, and
a write. Say 20ms total (Provided the disk is turning). I find that an
acceptable "price" to pay if I haven't written to a filesystem for say
a minute. (*) But if that's "too close" for you, feel free to set the
default timeout at half an hour.

I have a computer with 21G of "dead" harddisks. They would be nice to
have mounted, but I'm currently working on projects (driver
development) that crash the machine regularly, so I have to keep the
fstab from mounting them, otherwise I'd "dirty" them way too often...

			Roger.

(*) Moreover, it is highly likely that I'm writing to a mounted
filesystem. The system knows five to thirty seconds before the actual
first write that dirty buffers for that disk exist!

-- 
** R.E.Wolff en BitWizard.nl ** http://www.BitWizard.nl/ ** +31-15-2137555 **
*-- BitWizard writes Linux device drivers for any device you may have! --*
 "I didn't say it was your fault. I said I was going to blame it on you."

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo en vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/