> The outer nested join has the VALUES as the main loop, and the complicated
> join as the leaf. So, the complicated overlap-finding join gets run twice.
That's weird. What do you have as statistics target? Planner is incorrect
few orders of magnitude, so increasing it may help.
BTW: One of constraints is redundant l1.start <= l2.start implies l1.start
<= l2.end, so latter can be removed as for me.
> Oh, there's also the great big sort and unique, but I think I can get rid
> of that.
As far as I can see, duplicates will occur if and only if l1.start ==
l2.start && l1.end == l2.end.
That can be easily filtered by adding "where n=1 or l1.start != l2.start or
l1.end != l2.end" to outer select.
In response to
pgsql-performance by date
|Next:||From: Kevin Grittner||Date: 2009-03-31 21:35:59|
|Subject: Re: Strange behavior: pgbench and new Linux kernels|
|Previous:||From: Matthew Wakeling||Date: 2009-03-31 17:08:00|
|Subject: Re: Very specialised query|