Re: [PROPOSAL] Max recursion depth in WITH Queries (Common Table Expressions)

Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com> writes:
> 2015-10-28 14:33 GMT+01:00 Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>:
>> Also, there are already ways to constrain queries-gone-crazy; particularly
>> statement_timeout, which has the advantage that it works for other types
>> of badly-written queries not only this one.

> isn't the recursive limits much more a resource limit like work_mem etc?

Exceeding work_mem isn't generally supposed to result in an error --- it
causes, or should cause, the system to shift execution strategy so that
you get the same answer with less memory and more time consumption.

In any case, the question is what purpose this would serve that isn't
already covered perfectly well by existing features like
statement_timeout.

regards, tom lane