Thread: RE: Re: CRC

RE: Re: CRC

From
"Mikheev, Vadim"
Date:
> One thing we should look at before going with a 64-bit method is the
> extra storage space for the larger checksum.  We can clearly afford
> an extra 32 bits for a checksum on an 8K disk page, but if Vadim is
> envisioning checksumming each individual XLOG record then the extra
> space is more annoying.

We need in checksum for each record. But there is no problem with
64bit CRC: log record header is 8byte aligned, so CRC addition
will add 8bytes to header anyway. Is there any CRC64 code?

Vadim


Re: Re: CRC

From
Bruce Guenter
Date:
On Mon, Dec 11, 2000 at 10:09:01AM -0800, Mikheev, Vadim wrote:
> > One thing we should look at before going with a 64-bit method is the
> > extra storage space for the larger checksum.  We can clearly afford
> > an extra 32 bits for a checksum on an 8K disk page, but if Vadim is
> > envisioning checksumming each individual XLOG record then the extra
> > space is more annoying.
> We need in checksum for each record. But there is no problem with
> 64bit CRC: log record header is 8byte aligned, so CRC addition
> will add 8bytes to header anyway. Is there any CRC64 code?

All you need is a good 64-bit polynomial.  Unfortunately, I've been
unable to find one that's been analyzed to any amount.
--
Bruce Guenter <bruceg@em.ca>                       http://em.ca/~bruceg/