Re: [PERFORM] Inheritance, unique keys and performance

2008-01-13 Thread Julian Scarfe
From: "Tom Lane" <[EMAIL PROTECTED]> The problem you've got is that you need a nestloop-with-inner-indexscan plan type, where the inner side is an Append group (that is, an inheritance tree) --- and 8.1 doesn't know how to create such a plan. If you can update to 8.2 or later it should get bette

Re: [PERFORM] Inheritance, unique keys and performance

2008-01-13 Thread Julian Scarfe
Without the EXPLAIN ANALYZE output, nobody can say whether you have interpreted your performance problem correctly or not. Fair enough, Tom. superclass = "geonode", subclass = "airport", expensive_function = "gc_offroute". For this test, some_table is also "airport". There's also a coarse

[PERFORM] Inheritance, unique keys and performance

2008-01-12 Thread Julian Scarfe
Useing 8.1.9 I'm running into some performance issues with inheritance. I've abstracted the situation below because otherwise there's lots of spurious stuff, but I'll gladly provide the real EXPLAIN ANALYZE output if necessary. I have a superclass and a dozen subclasses, of which I've picked o