Re: Large Scale Aggregation (HashAgg Enhancement)

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Simon Riggs <simon(at)2ndquadrant(dot)com>
Cc: Rod Taylor <pg(at)rbt(dot)ca>, PostgreSQL Development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Large Scale Aggregation (HashAgg Enhancement)
Date: 2006-01-19 23:38:11
Message-ID: 10671.1137713891@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Simon Riggs <simon(at)2ndquadrant(dot)com> writes:
> This seems to lead to a super-geometric progression in the number of
> files required,

But we double the number of batches at each step, so there are going to
be at most 20 or so levels, and that's only assuming a *horridly* wrong
initial guess by the planner. In practice I think it's reasonable to
assume at most a couple rounds of doubling. If you have more than that,
the extra data-shuffling is going to exhaust your patience anyway.

regards, tom lane

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Tom Lane 2006-01-20 00:16:03 Re: BuildFarm: Do we need another FreeBSD/amd64 member?
Previous Message Tom Lane 2006-01-19 23:26:09 Re: Bug: random() can return 1.0