On 4/13/06, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> "patrick keshishian" <pkeshish@gmail.com> writes:
> > With these settings and running:
> > pg_restore -vaOd dbname dbname.DUMP
>
> If you had mentioned you were using random nondefault switches, we'd
Random?
With all due respect, I did.
I specified the PostgreSQL version of the pg_dump source
server. I specified the version of my dev PostgreSQL server.
I provided specific information about which postgresql.conf
entries I had changed and to what specific values they were
changed to.
I pasted the _exact_ command used (including so called
"random nondefault switches") to do the dump and the
exact command used (again, with said "random nondefault
switches") to restore from the dump'ed data.
I believe I tried my best to be as thorough as possible with
my post(s).
archived at: http://archives.postgresql.org/pgsql-performance/2006-04/msg00287.php
> have told you not to. -a in particular is a horrid idea performancewise
> --- a standard schema-plus-data restore goes way faster because it's
> doing index builds and foreign key checks wholesale instead of
> incrementally.
Duly noted. Option "-a" bad.
> > Is this because the -c option drops all foreign keys and
> > so the restore goes faster? Should this be the preferred,
> > recommended and documented method to run pg_restore?
>
> It is documented in recent versions of the documentation: see
> http://www.postgresql.org/docs/8.1/static/populate.html
> particularly the last section.
As a general rule of thumb, I have always assumed,
documentation for any software, from one major version
to another, would not necessarily apply cross revisions
(e.g., version 7.4 vs 8.1).
But thanks for your time and help,
--patrick