Rod Taylor wrote:
> On Tue, 2006-06-06 at 16:04 +0000, Christopher Browne wrote:
>   
>> 2.  There *might* be some cleverness about closing out transactions on
>> the provider so that the subscription process for 4 tables would
>> represent 4 transactions on the provider.
>>     
>
> I'm not familiar enough with the process. What happens on the provider
> after the set is copied and the indexes are built on the subscriber?
>
> Is it as simple as removing the entry from sl_event or something
> substantially more complex?
>   
The "next part" that is tricky is that the first SYNC after the copy
needs to figure out which updates in sl_log_1/2 were not included with
the COPY.  If you have 17 transactions for 17 tables, that means having
to evaluate SYNCs for each table against 17 different bases.

As it stands, by doing it all as one transaction, on the provider, you
have just 1 transaction to reference to figure out what was and wasn't
applied when the COPY finished.

Breaking it out to 17 transactions for 17 tables would seem *plausible*,
but certainly more complicated.
_______________________________________________
Slony1-general mailing list
[email protected]
http://gborg.postgresql.org/mailman/listinfo/slony1-general

Reply via email to