(http://www3.sk.postgresql.org/docs/faqs/FAQ.html):
OIDs are stored as 4-byte integers, and will overflow at 4 billion. No
one has reported this ever happening, and we plan to have the limit
removed before anyone does.
That comment in the FAQ seems quite out-of-date.
What will actually happen is that the OID generator will wrap around. This will not bother Postgres particularly, but you may start having occasional transaction failures due to duplicate OIDs --- for example, I believe lo_create will fail if the OID it selects already exists in pg_largeobject.
Pardon my incredulity, but doesn't that seem like a bug? Or at least a limitation? Does this mean that the effective useful lifetime of pg_largeobject is only as long as it takes to wrap around, after which you *must* dump and reload to prevent problems like this?
(I realize this is pretty much the same issue as having a sequence number on a table, but if I'm interpreting all this correctly, the OID wrap-around is going to occur a lot sooner than my table sequence number wrap-around.)
-- Jeff Boes vox 269.226.9550 ext 24 Database Engineer fax 269.349.9076 Nexcerpt, Inc. http://www.nexcerpt.com ...Nexcerpt... Extend your Expertise
---------------------------(end of broadcast)--------------------------- TIP 7: don't forget to increase your free space map settings
