Brad Nicholson wrote:
> > When you say "corrupted", I assume you mean you have remove the _old
> > suffixes to restart your 8.3 instance, right? I hope that is the only
> > corruption issue --- please confirm.
>
> Unfortunately no - when I try and start the old version, I get:
>
> pg_ctl -D pgsql83/ start
> postgres: could not find the database system
> Expected to find it in the directory "/opt/DATA/pgsql83",
> but could not open file "/opt/DATA/pgsql83/global/pg_control": No such
> file or directory
>
> I am specifying both new and old paths to pg_migrator, as I don't have
> my data directories in standard locations.
>
> Test case is pretty straight forward.
>
> Create an 8.3 instance
> Add dblink
> follow through the pg_migrator upgrade path
> try and start your 8.3 instance after the failure
Uh, I assume you read all of the INSTALL instructions, including the
last item:
10. Reverting to old cluster
...
--> If you ran pg_migrator _without_ --link or did not start the new server,the old cluster was not modified except
thatan ".old" suffix wasappended to $PGDATA/global/pg_control and tablespaces directories. Toreuse the old cluster,
removethe tablespace directories created by thenew cluster and remove the ".old" suffix from the old cluster
tablespacedirectorynames and $PGDATA/global/pg_control; then you can restart theold cluster.
I just modified the arrow text to be clearer. The rename of
pg_controldata is done to prevent accidental starting of the old cluster
in case of migration success.
-- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB
http://enterprisedb.com
+ If your life is a hard drive, Christ can be your backup. +