The following bug has been logged online:
Bug reference: 3692
Logged by: Bill Moran
Email address: wmoran@collaborativefusion.com
PostgreSQL version: 8.2.5
Operating system: FreeBSD
Description: Conflicting create table statements throw unexpected
error
Details:
(also occurs on 8.1.10)
Issuing a statement like:
CREATE TABLE table2 AS SELECT * FROM table1;
simultaneously in two separate sessions should result in an error like
"ERROR: relation "table2" already exists" (in one or the other of the
sessions, depending on the exact timing of things).
However, if table1 has enough rows that the command takes a while to execute
(a few seconds seems to be all it takes) the error is far more cryptic:
ERROR: duplicate key violates unique constraint "pg_type_typname_nsp_index"
It seems to me that there's some sort of race condition that if the second
command starts before the first has completed, the backend doesn't really
understand what went wrong.
For a front end, this is tough to parse. A "relation exists" error on a
table should probably be 42P07, but the duplicate key violation results in
23505, which means a front end will likely behave incorrectly.