[Sugar-devel] Additional data for corrupted SoaS

[Dave Bauer]

On Tue, Aug 4, 2009 at 9:11 AM, Sascha
Silbe<sascha-ml-ui-sugar-devel at silbe.org> wrote:
> On Tue, Aug 04, 2009 at 01:52:31PM +0100, Martin Dengler wrote:
>
>>> Some discussion on irc of possible solutions [to ext3 problems]:
>>
>> We should be using ext2, not ext3, in the future so that we don't
>> have a journal to corrupt.
>



Sorry that should say Datastore, not journal, as journal referred to
the user interface to the Datastore.

Dave

> That's no solution. You'll just corrupt the rest of the filesystem instead,
> with no information about what parts of the filesystem might be damaged.
> For simple "media goes offline" / power outage cases, journal=data should be
> enough to preserve the integrity of the device. For altered-data cases (that
> I sure hope won't happen with regular USB sticks!), you'll need something
> that adds redundancy, i.e. some kind of single-device RAID5. I don't know of
> anything in the Linux kernel that would provide this kind of thing (for
> regular block devices, not raw NAND access).
>
> CU Sascha
>
> --
> http://sascha.silbe.org/
> http://www.infra-silbe.de/
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.9 (GNU/Linux)
>
> iQEcBAEBAgAGBQJKeDONAAoJELpz82VMF3DaoRoIAJU2H/tBfPYTquLRa23B8ZN5
> jrFQohJiW9hOmVHxJkt+ViZIM7Hz1bNKSk1XyT5W6pkKVScJQByozoBqNUo1KmpI
> U2SEUWvDmyNfoU0YRN6h9Rh/L/E5jOotz31/WOxmt/JmhGHd1/ZuF7ZQotvndyhu
> 511mTl+tE9IQZVehjDMY9Q/VGmLRV7s7s2vQzAK4qEs+s7AMPM6Bmab9aoA/5LFa
> imbWmVRdqu08xDocBp7bS/s67ZWm/U72Y2QeJDkmBiYpCA6ljMZu9xgfVEOFps9A
> 69lYtcClrKY6sFvPqy/242oidrZmoeM/VlJThU2pR8q1Req1UmgIncssqiqjfzs=
> =VyQl
> -----END PGP SIGNATURE-----
>
>



-- 
Dave Bauer
dave at solutiongrove.com
http://www.solutiongrove.com