On November 19, 2004 01:44 pm, you wrote:
> On November 19, 2004 12:27 pm, Christopher Browne wrote:
> > Darcy Buskermolen <darcy at wavefire.com> writes:
> > > On November 18, 2004 07:48 am, Christopher Browne wrote:
> > >>I have the following as somewhat "version-unaware" upgrade scheme
> > >>which can cope with the case of needing to add a column to a table:
> > >>
> > >> create or replace function @NAMESPACE at .add_missing_table_field (text,
> > >> text, text, text) returns bool as '
> > It's something of a "stub" function, at this point; the notion is to
> > have a function there for upgrade activities to accrete into.
> >
> > Further, it's a bit of a "trial balloon;" if it seems sensible, then
> > it's sensible to continue.  If it's not, then it should disappear...
>
> it looks to be senseable to me, Just a matter of ironing out the details of
> how we want to handle the actual upgrade in slonik beyond just sending a
> notify _cluster.Restart.  Chris, what were your thoughts on how you figured
> this should be handled.  At the moment I'm leaning towards a function in
> slony_funcs.sql that you just serialize the placement of  perform
> upgrade_foo into?

Further to this discusion I have prepared a bit more detail on what I'm 
envisioning in the upgrade method. 

http://www.dbitech.ca/slony/slonyupgrade.diff  any comments or theory 
debunking would be most welcomed.


-- 
Darcy Buskermolen
Wavefire Technologies Corp.
ph: 250.717.0200
fx:  250.763.1759
http://www.wavefire.com