On Wed, 18 Jun 2008, Joshua D. Drake wrote:
This is what I suggest: download 8.1.13: unpack; then:
Done.
./configure --prefix=/tmp/pg813; make install cd /tmp/pg813 bin/pg_ctl -D /usr4/pgsql_old/data start cd / /usr/bin/pg_dumpall -U <user> > mydatabase.sql
Modified above a bit. I used /usr3/pg813, needed to su to postgres to start the daemon, and cd to /usr4/pgsql_old/ to run /usr/bin/pg_dumpall. That file now exits in /usr4/postgres-backups and is larger than yesterday's attempt.
At this point you will have an 8.3 dump of your 8.1 data.
Looks good to me.
Then you can reinitialize a new cluster with initdb here: /var/lib/pgsql/data (you will have to remove the old one)
As user postgres, I cleaned out /var/lib/pgsql/data/* and re-initialized. I had to specify -E UTF8 because 8.3.3 is not finding en_US as a valid locale coding.
Then restore as normal using psql -U postgres < mydatabase.sql
Can't get here. As user postgres, I shut down the 8.1.13 process. However, I cannot start the 8.3.3 daemon running. I tried as user postgres with the pg_ctl command, and as root running the formerly-working '/etc/rc.d/rc.postgresql start' command. What I see with the latter is: [root@salmo /etc/rc.d]# ./rc.postgresql start Starting PostgreSQL 18162 PostgreSQL daemon already running Warning: Missing pid file /var/lib/pgsql/data/postmaster.pid But, there is no postmaster process running despite having the process ID of 18162 shown. Sure enough, the pid file is not present, nor is process 18162 there. Much closer, Josh, but not quite there yet. Many thanks, Rich -- Richard B. Shepard, Ph.D. | Integrity Credibility Applied Ecosystem Services, Inc. | Innovation <http://www.appl-ecosys.com> Voice: 503-667-4517 Fax: 503-667-8863
