Re: foreign key locks - EvalPlanQual interactions

On 2012-11-27 12:07:34 -0300, Alvaro Herrera wrote:
> Alvaro Herrera wrote:
> > Here's version 24.
>
> Old review emails still contains some items that didn't lead to any
> changes.  I tried to keep close track of those.  To that list I add a
> couple of things of my own.  Here they are, for those following along.
> I welcome suggestions.
>
> - EvalPlanQual and ExecLockRows maybe need a different overall locking
>   strategy.  Right now follow_updates=false is used to avoid deadlocks.

I think this really might need some work. Afaics the EPQ code now needs
to actually determine what locklevel needs to be passed to
EvalPlanQualFetch via EvalPlanQual otherwise some of the locking issues
that triggered all this remain. That sucks a bit from a modularity
perspective, but I don't see another way.

Greetings,

Andres

-- Andres Freund                       http://www.2ndQuadrant.com/PostgreSQL Development, 24x7 Support, Training &
Services