Heikki Linnakangas <hlinnaka@iki.fi> writes:
> On 09/24/2016 02:34 PM, Peter Geoghegan wrote:
>> I go on to explain how this patch represents a partial solution to
>> that [1]. That's what I mean by "useful practical consequences". As I
>> say in [1], I think we could even get a full solution, if we applied
>> this patch and *also* made the ordering in which the relcache returns
>> a list of index OIDs more useful (it should still be based on
>> something stable, to avoid deadlocks, but more than just OID order.
>> Instead, relcache.c can sort indexes such that we insert into primary
>> keys first, then unique indexes, then all other indexes. This also
>> avoids bloat if there is a unique violation, by getting unique indexes
>> out of the way first during ExecInsert()).
> Hmm, yeah, that'd be nice to fix. I'd like to see a patch specifically
> to fix that. I'm not convinced that we need all the complications of
> this patch, to get that fixed. (In particular, indexam's still wouldn't
> need to care about the different between CHECK_UNIQUE_PARTIAL and
> CHECK_UNIQUE_SPECULATIVE, I think.)
I can see the value of processing unique indexes before non-unique ones.
I'm pretty suspicious of trying to prioritize primary keys first, though,
because (a) it's not clear why bother, and (b) PG is a tad squishy about
whether an index is a primary key or not, so that I'd be worried about
different backends sometimes choosing different orders. I'd simplify
this to "uniques in OID order then non-uniques in OID order".
regards, tom lane