> Did we decide against LAZY? Seems we have a number of people
> concerned about vacuum downtime, and I can see this as a win
> for them. If they don't specify LAZY, the code is not run.
First sorry that I wasn't able to deal with vlazy earlier.
Now I have one more open item for 7.1 - restoring index structure
at runtime (add tuple to parent page for aborted split op,
create root page if no one exists). I'll try to deal with both
items (in any case vlazy will be ported to 7.1, as required
by contract).
As for reported problem: I just looked at Denis' tgz and
found only table, index and sequence(?) data files - I would
need in schema definitions, pg_log and pg_variable files
as well. Denis? Also, when these copies were made -
before/after unsuccessful vacuum+lazy?
Vadim