[Pgcluster-general] Best practice for bringing up clustered DBs

[Leif Madsen]

Morning all!

I've been using pgcluster in my development environment lightly for
about a month now, and really enjoying the functionality of it. Great
work.

Unless I've totally missed the boat, I've not found much documentation
on using pgcluster, but was able to manage getting it up and running
from the INSTALL_PGCLUSTER text file included.

My question is basically this:

Once all the configuration files are setup for pgreplicate and
cluster, what is the best way to be bringing the databases up for the
first time. From my experience, it seems that I want to do something
like:

./bin/pgreplicate -D /usr/local/pgsql/data

./bin/pg_ctl start -D /usr/local/pgsql/data (Master DB?)

./bin/psql < my_db_dump.txt

Then, on the other clustered DBs, to do something like:

./bin/pg_ctl start -D /usr/local/pgsql/data -o "-U"

I've not had any luck with -R so far, and -r actually returns an error
for me. Perhaps I'm not using those flags in the correct scenarios?
(If someone would like to expand upon when to use -r, -R, and -U,
that'd be very helpful.)

Any additional information for "first time" clustered DB start up, or
database recovery would be grand. I do plan on writing some basic
documentation for pgcluster in the near future once I have a bit more
experience with it. So hopefully your answers won't be all for not :)

Thanks!

-- 
Leif Madsen.
http://www.leifmadsen.com
http://www.oreilly.com/catalog/asterisk