Vincent Chen wrote:

>Hi, all
>
>I just setup a postgresql cluster and some questions come to me:
>
>1. slon running on foreground and keep showing logs, how can I configure slon
>run on background as daemon?
>  
>
Typically, we write wrapper scripts to start the slon.

>2. my test environment has 1 master and 1 slave node. if I need to add another
>slave node, how can I do that?
>  
>
There's no such thing as a "master node" or a "slave node." They are all
just nodes.

If you need to add another node, then...
1. Copy the schema over to the new database,
2. Use STORE NODE to initialize the new database as a Slony-I node
3. Use SUBSCRIBE SET to subscribe the new node to your replication set(s).

>3. mysql also has replication ability. what's the difference between these 2
>DBMS cluster solution?
>  
>
I haven't used the MySQL replication system so can't comment on it
meaningfully.

But when cooking, I find that starting with good quality ingredients
turns out a lot better than grabbing ingredients of dubious integrity.
The latter may be satisfactory for generating some sort of stew to which
one adds heavy spices so you can't taste the ingredients.

I'd generally rather start cooking my database applications with
PostgreSQL...