> > Could the index scan be made to
> > handle cases where the index tuple it was stopped on is gone?
>
> Don't see how. With no equal keys, you could test each tuple you scan
> over to see if it's > the expected key; but that would slow things down
> tremendously I fear. In any case it fails completely when there are
> equal keys, since you could not tell where in a run of equal keys to
> resume scanning. You really have to find the exact index tuple you
> stopped on, AFAICS.
Won't it still point to the same heap page and slot ? That additional info
should be sufficient to find the exact index tuple.
And it usually won't be that far away, no ?
Andreas