Gavin Scott <gavin@pokerpages.com> writes:
> Just happened again this morning. It turns out this looks like an
> openssl problem, not a postgresql one:
Hmm. It'd be worth your while to update to Postgres 7.2. In 7.2 the
postmaster forks a child process before it starts to do anything with
the connection, so hangups in SSL can't lock the whole postmaster, only
the one connection. Or that's the theory anyway.
regards, tom lane