I read the discussion at
From what I can understand, going from/to unlogged to/from logged in
the wal_level == minimal case is not too complicated.
Suppose I try to write a patch that allows
ALTER TABLE tablename SET LOGGED (or UNLOGGED)
(proper sql wording to be discussed...)
only in the wal_level == minimal case: would it be accepted as a
"first step"? Or rejected because it doesn't allow it in the other
From what I got in the discussion, handling the other wal_level cases
can be very complicated (example: the issues in case "we *crash*
without writing an abort record").
pgsql-hackers by date
|Next:||From: Noah Misch||Date: 2011-04-08 11:08:13|
|Subject: Re: pg_upgrade bug found!|
|Previous:||From: Brendan Jurd||Date: 2011-04-08 09:00:22|
|Subject: Re: Failed assert ((data - start) == data_size) in heaptuple.c|