Bruce Momjian <bruce@xxxxxxxxxx> writes: > My guess is that having a very long fuzzy checkpoint time means that > when you fsync you are fsync'ing lots of data, both your checkpoint data > and other writes performed by backends. Hmm, could be ... although that would imply that shared_buffers should be kicked up some more, so the backends aren't doing so many writes for themselves. regards, tom lane -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general