Forum OpenACS Development: Re: Using slony-1 / postgres for a openacs based site.

Collapse
Posted by Bjoern Kiesbye on
Hello, thanks,

I have a test cluster for oacs up and running, till now every think works fine.

Here is a collection of scripts.
  • a sql script to alter (add primary keys) tables from acs-* packages which don't have a primary key. a slonik sh script which initializes a cluster . Creates slony sets (~ 30) and adds the tables / sequences (~200) of the acs-* packages to the sets.
  • And several helper scripts to start/stop replication , subscribing / unsubscribing nodes , cluster wide shema changes, and so on.
  • As well there is a installation / configuration document of slony / pgpoolsII / postgres / aolserver, its not yet a real documentation, more extensive notes I made during setup. (the installation of slony is not in there , but a lot about configuring slony ).

I hope they will be helpfully .

The system I need this for is not using acs extensively, so I cant say for sure that every think is working with the given configuration, but creation of acs_objects , adding items to the content repository or creating news items, worked fine.
The synchronization is faster than I expected, on the test system subscribernodes where synchronized about 1 sec. after a update or insert changed the Masterdb.

Hope to here from your progress Eduardo.
Collapse
Posted by Bjoern Kiesbye on
Hi,

I was just doing an installation of slony, and added the installation steps to the installation / configuration document.