On Friday 04 March 2005 14:24, Vivek Khera wrote:
> On Mar 4, 2005, at 5:13 PM, Jan Wieck wrote:
> >> -rw-r--r--  1 root  wheel  2916941855 Mar  3 04:38 mm.03-Mar-2005.dump
> >> -rw-r--r--  1 root  wheel  2359870138 Mar  4 11:51 mm.04-Mar-2005.dump
> >> the March 2 and 3 dumps were from the origin, and march 4 from the
> >> replica.  This is a "-Fc" format pg_dump of the entire database.
> >> And yes, I'm still dumping the schema from the origin also :-)
> >
> > But you do dump the schema with the data anyway - which is useless and
> > corrupted on the subscriber, so you really want to have a data-only
> > dump from it. A data only dump from origin and subscriber can only
> > differ in the data contained in the sl_log_1 and sl_event tables.
>
> yeah, i can just ignore the schema in that dump, too!
>
> but how on earth can they differ by 500Mb+ *compressed*?
>
> all the data *seems* to be there.  i guess i really need to
> re-double-triple-quadruple check it...

Slow running backup with lots of churn could result in millions of rows in 
sl_log/event tables.  That could explain your difference right there.


>
> Vivek Khera, Ph.D.
> +1-301-869-4449 x806
>
> _______________________________________________
> Slony1-general mailing list
> Slony1-general at gborg.postgresql.org
> http://gborg.postgresql.org/mailman/listinfo/slony1-general