Thanks raghu ram for your immediate response,

i will try this path too along with = my team and i will come back if any issues.

Thanks

Senthil

From: raghu = ram [mailto:raghuchennuru@gmail.com]
Sent: Friday, April 01, = 2011 8:57 PM
To: Senthil Kumar G
Cc: = pgsql-admin@postgresql.org
Subject: Re: [ADMIN] DB = Import Error...

On Fri, Apr 1, 2011 at 8:34 PM, Senthil Kumar G <senthil@zoniac.com> = wrote:

Hi<= /p>

I am upgrading my staging environment from postgresql 8.2.3 to = 9.0.3.

I was able to do successfully.

But, when i try to import the database which is created in 8.2.0 version to 9.0.3 = version environment, i am getting following error.

ERROR: constraint "xxxx” for relation "xxxx” already = exist

Based on above error message it looks like the constraint for relation already = exists.

Could you please follow below steps to upgrade = from older version to new version::

Step 1: = Perform the global dump on old cluster [ i.e PostgreSQL 8.2.3] using "pg_dumpall" binary of new PostgreSQL = 9.0.

/opt/PostgreSQL/9.0/bin/pg_dumpall -d postgres -U postgres -p 5432 -g > /tmp/globaldump_oldpg823.sql

NOTE: = It Dumps only global objects i.e roles,users and tablespaces,no databases.

Assuming old cluster running on the 5432 port = number.

Step 2: = Take the dump of the database in compressed format using new = version of pg_dump binary.

Syntax: /opt/PostgreSQL/9.0/bin/pg_dump -Fc -v = -p 5432 -f <dump file = location> <database name>

nohup /opt/PostgreSQL/9.0/bin/pg_dump -Fc -v -p = 5432 -f /tmp/demo.dmp demo>> /tmp/dump.log 2>> /tmp/dump.log = &

-P =3D> port number of old = cluster

-Fc =3D> compressed format

-v =3D> verbose output

-f =3D> location of dump file to store and dump file = name

NOTE: = Assuming old cluster running on the 5432 port = number

Step 3: = Restore the global dump on new cluster of = PostgreSQL 9.0.2

/opt/PostgreSQL/9.0/bin/psql -d postgres -U = postgres -p 5433 -f /tmp/globaldump_oldpg832.sql

NOTE: = Assuming new cluster running on the 5433 port = number

Step 4: = Restore the compressed dump file using new = version of pg_restore binary with parallel = restore operation.

nohup /opt/PostgreSQL/9.0/bin/pg_restore -Fc -v = -p 5433 -d demo -j 8 /tmp/demo.dmp >>/tmp/restore1.log = 2>>/tmp/restore1.log &

NOTE: = Assuming new cluster running on the 5433 port = number

--Raghu Ram

when i query it was not found.

What could be the reason? What should i do to resolve this? =

Appreciate your quick answer.

Thanks & Regards

Senthil