Bruce Momjian <bruce(at)momjian(dot)us> writes:
> Magnus reported that a customer with a million tables was finding
> pg_upgrade slow.

You sure there's not an O(N^2) issue in there somewhere?

> I don't see anything unsafe about having pg_upgrade use
> synchronous_commit=off.

No objection, but this seems unlikely to be better than linear speedup,
with a not-terribly-large constant factor.

BTW, does pg_upgrade run pg_restore in --single-transaction mode?
That would probably make synchronous_commit moot, at least for that
step.

			regards, tom lane

In response to

• Pg_upgrade speed for many tables at 2012-11-05 20:08:17 from Bruce Momjian

Responses

• Re: Pg_upgrade speed for many tables at 2012-11-05 20:23:04 from Magnus Hagander
• Re: Pg_upgrade speed for many tables at 2012-11-05 23:15:30 from Jeff Janes

pgsql-hackers by date

Next:	From: Robert Haas	Date: 2012-11-05 20:19:01
	Subject: Re: Limiting the number of parameterized indexpaths created
Previous:	From: Bruce Momjian	Date: 2012-11-05 20:08:17
	Subject: Pg_upgrade speed for many tables