http://www.slony.info/bugzilla/show_bug.cgi?id=181

--- Comment #6 from Christopher Browne <cbbrowne at ca.afilias.info> 2011-01-26 12:03:01 PST ---
I noticed some error conditions on this where we should declare explicitly what
we intend to happen...

The addition of tables will fail, for ALL the tables, if any of the following
problems are encountered:

 a) A table does not have a primary key.  Note that a candidate primary key
    is not good enough; since we do not have an option here to specify 
    candidate primary keys, the use of this requires an explicit primary
    key.

 b) A table is already replicated.

Both of those conditions seem apropos to me; if you want to use regular
expressions to specify a set of tables, you're going to have to have the tables
suitably defined so we don't have to jump thru too many hoops to ascertain
replication configuration for them.

And by the way, this should be added to the man pages for SET ADD
TABLE/SEQUENCE...

I suggest the following positive and negative tests; they probably should get
implemented in the "cluster test" framework in the other branch...

1.  Put a series of nice, replicable tables in one schema...

Tables complete with primary keys, and regular names so they're readily
matched.

Should work, no problems.

2.  Negative test:  Redo 1.  

The attempt to re-add them should fail.  "Kaboom!"

3.  Have a node in the preamble that hasn't had STORE NODE called on it yet.

Run against a set of tables not added yet.

4.  Have a node that's dead.

-- 
Configure bugmail: http://www.slony.info/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.