pg_upgrade cleanup

On Mon, Apr 25, 2011 at 8:19 PM, Bruce Momjian <bruce@momjian.us> wrote:

The attached applied patch improves pg_upgrade by:  avoiding one
start/stop of the postmaster;  using the -w (wait) flag for pg_ctl
start/stop;  removing the unused "quiet" flag in the functions for
starting/stopping the postmaster.

It is well past time to stop tweaking pg_upgrade for 9.1. Feature
freeze was three months ago, and the risk of introducing new bugs at
this point surely outweighs any gain we might get from the changes.
We should only be fixing *bugs* at this point.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Robert Haas wrote:

On Mon, Apr 25, 2011 at 8:19 PM, Bruce Momjian <bruce@momjian.us> wrote:

The attached applied patch improves pg_upgrade by: ?avoiding one
start/stop of the postmaster; ?using the -w (wait) flag for pg_ctl
start/stop; ?removing the unused "quiet" flag in the functions for
starting/stopping the postmaster.

It is well past time to stop tweaking pg_upgrade for 9.1. Feature
freeze was three months ago, and the risk of introducing new bugs at
this point surely outweighs any gain we might get from the changes.
We should only be fixing *bugs* at this point.

Someone at PG East complained pg_upgrade wasn't fast enough for his
usage, so I tried to speed it up. (I guess it was a bug for him.) I
think not using -w for pg_ctl could be a potential bug because we wack
around the files underneath the postmaster when we think it is shut
down. The "quiet" change was just a cleanup.

--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com

+ It's impossible for everything to be true. +