Synopsis: 8-table join with one "WHERE foo IN (...)" condition; works OK with
fewer
than 5 items in the IN list, but at N=5, the planner starts using a compound
index
for the first time that completely kills performance (5-6 minutes versus 0-12
seconds).
I'm interested in learning what plays a
Sorry for the delay; back on this, and thanks for the response.
On Mar 11, 2014, at 6:23 PM, Tom Lane wrote:
> "Murphy, Kevin" writes:
>> Synopsis: 8-table join with one "WHERE foo IN (...)" condition; works OK
>> with fewer
>> than 5 items in the IN l
