The TeamForge upgrade utility upgrades PostgreSQL for you automatically when you run
the prepare-environment.sh script. However, when you upgrade to
TeamForge 6.1 from TeamForge 5.4, patch 1, it is possible for the automatic PostreSQL
upgrade to fail. If this happens, you can do the PostgreSQL upgrade yourself.
The error, if it occurs, will look like this:
ERROR: postgres upgrade failed.
Exiting due to fatal error.
-
Create a database dump.
Note: If you have already made the dump, skip this step.
- su - postgres -c /usr/bin/pg_dumpall >
/tmp/dumppath/
-
Upgrade PostgreSQL.
-
Ensure that PostgreSQL is running.
- /etc/init.d/postgresql-9.0 start
-
Upgrade the PostgreSQL packages.
- yum install postgresql postgresql-server
postgresql-docs
-
If you get this error:
-
%postun(postgresql-server-8.3.8-1PGDG.rhel5.x86_64)
scriptlet failed, exit status 1
then run this command:
- yum erase
postgresql-server-8.3.8-1PGDG.rhel5
-
Move the old pgsql directory out of the way.
- mv /var/lib/pgsql/9.0/ /var/lib/pgsql/9.0_old
-
Rerun the prepare-environment.sh command. When it runs
successfully, you can run install.sh and
migrate.py.
Important: After you run the installer and before you run the
migration command, you must reload the PostgreSQL data dump:
- su - postgres -c /usr/bin/psql <
/tmp/dumppath/<name>.dmp