On Thu, Jun 08, 2006 at 11:55:02AM -0400, Tom Lane wrote:
> "Jim C. Nasby" <jnasby@pervasive.com> writes:
> > On Thu, Jun 08, 2006 at 11:18:02AM +0200, Ludek Finstrle wrote:
> >> I read this value in TOAST section. Is my opinion correct?
>
> > From http://www.postgresql.org/docs/8.1/interactive/storage-toast.html:
> > "TOAST usurps the high-order two bits of the varlena length word,
> > thereby limiting the logical size of any value of a TOAST-able data type
> > to 1Gb (2^30 - 1 bytes)."
>
> > There was a proposal made some time ago to allow for a variable-length
> > length word format, where one of the bits in each word would specify
> > that there was an additional length word.
>
> Hm, I don't remember that. It seems rather pointless, as I'm quite sure
> that the *practical* limit is a great deal less than 1Gb. Has anyone
> done any performance testing of GB-sized toasted values?
Given how toasted data is currently stored, you're probably correct. If
it was switched to a binary format that didn't have all the table/tuple
overheard (which I seem to recall a discussion about), it could be a
very different story.
--
Jim C. Nasby, Sr. Engineering Consultant jnasby@pervasive.com
Pervasive Software http://pervasive.com work: 512-231-6117
vcard: http://jim.nasby.net/pervasive.vcf cell: 512-569-9461