Teodor Sigaev wrote:
>> Right now, if an index entry points to a dead tuple, we set a bit in
>> the index so future lookups do not access the heap. We could set a
>> bit for block index entries that point to a page that has no live
>> rows, and
>> have vacuum remove the index entry later.
>
> GIN don't support this feature...
I'm only talking about B-trees at this stage. ISTM that you could do the
same thing with hash indexes, but I haven't given it much thought.
Anyway, I think you'd usually want to use bitmap scans with a Block
B-tree, unless you need sorted output. And bitmap scans don't set the
LP_DELETE flag either. We might want to do something about that.
--
Heikki Linnakangas
EnterpriseDB http://www.enterprisedb.com
