>> Yeah -- why is LOCK SEQUENCE foo_seq not allowed? Seems a simple thing
>> to have.
>
> I don't see any particular reason to continue to disallow it, but does
> that actually represent a workable solution path for pgpool? Switching
> over to that would fail on older servers.
pgpool will provide following method for older version of PostgreSQL.
> Probably creating a "secret" relation and acquire table locking
> on it is the way to go. This is essentially a dirty alternative for
> sequence table locking.
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese: http://www.sraoss.co.jp