Thread: How to use PG_DUMP?

How to use PG_DUMP?

From
Andre Lopes
Date:
Hi,
 
I'am having trouble using PG_DUMP. The problem is the following, I have made some minor changes to my database, I have added one table.
 
I have generated the CREATE statements using a modeling tool, and I have created another database to test the changes. My problem is the order of the INSERT statements generated by PG_DUMP [-a -d], causing errors because of the order of the INSERTS.
 
I have tested PG_DUMP [-a] that creates the COPY statements, but no data is added to the tables.
 
How can I solve this?
 
 
Best Regards, 

Re: How to use PG_DUMP?

From
Tom Lane
Date:
Andre Lopes <lopes80andre@gmail.com> writes:
> I have generated the CREATE statements using a modeling tool, and I have
> created another database to test the changes. My problem is the order of the
> INSERT statements generated by PG_DUMP [-a -d], causing errors because of
> the order of the INSERTS.

If you are talking about foreign key constraint failures, the best way
is to use a regular dump (not -a).  pg_dump will order the data and
constraint creations properly then.

            regards, tom lane

Re: How to use PG_DUMP?

From
Andre Lopes
Date:
Hi,

I have tested with "pg_dump -u -p 5432 -d  -f c:\test.sql mydatabase" but the order of the INSERTS it is not the correct. What PG_DUMP does is to add the CONSTRAINTS after doing the INSERTS. There is a way to have the correct order of the INSERTS?

I have tested the COPY but does not insert nothing.

What can I do?


Best Regards,


On Sun, Jan 24, 2010 at 5:22 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Andre Lopes <lopes80andre@gmail.com> writes:
> I have generated the CREATE statements using a modeling tool, and I have
> created another database to test the changes. My problem is the order of the
> INSERT statements generated by PG_DUMP [-a -d], causing errors because of
> the order of the INSERTS.

If you are talking about foreign key constraint failures, the best way
is to use a regular dump (not -a).  pg_dump will order the data and
constraint creations properly then.

                       regards, tom lane