Looking for speakers on High Performance Computing at GTALUG

[Lennart Sorensen]

On Mon, Jan 25, 2010 at 06:37:44PM -0500, Christopher Browne wrote:
> On Mon, Jan 25, 2010 at 5:38 PM, Lennart Sorensen
> <lsorense-1wCw9BSqJbv44Nm34jS7GywD8/FfD2ys at public.gmane.org> wrote:
> > I know that's what he meant.  I still was wondering what a distrusted
> > filesystem would look like.  Perhaps NFS or CIFS would come to mind. :)
> 
> There has been some ongoing discussion of this with respect to PostgreSQL.
> 
> This has been on the ToDo list for some time, with discussion as
> recent as last month.
> 
> Add optional CRC checksum to heap and index pages
> One difficulty is how to prevent hint bit changes from affecting the
> computed CRC checksum.
> http://archives.postgresql.org/message-id/19934.1226601952%40sss.pgh.pa.us
> http://archives.postgresql.org/pgsql-hackers/2008-10/msg00002.php
> http://archives.postgresql.org/pgsql-hackers/2008-10/msg01028.php
> http://archives.postgresql.org/pgsql-hackers/2008-11/msg00524.php
> http://archives.postgresql.org/pgsql-hackers/2008-12/msg01101.php
> http://archives.postgresql.org/pgsql-hackers/2009-12/msg00011.php
> 
> The discussion hasn't settled into offering much of a way to
> *implement* checksums to get real use out of them yet.  Indeed, it has
> had a tendency towards circular arguments, from one perspective or
> another:
> http://archives.postgresql.org/pgsql-hackers/2009-12/msg00042.php
> 
> I'm not quite certain I properly understand what a "Torn Page" is :-0

I guess this is why Oracle is sponsoring the development of BTRFS.
A filesystem with block checksuming would make the database much more
likely to trust the storage device.  Online fsck would also help uptimes.

-- 
Len Sorensen
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists