Обсуждение: Duplicate table name within the sme schema
Cyril DE MURCIA <cyril.de-murcia@dcnsgroup.com> writes: > We use postgresql 7.3.2 under Linux Red Hat 9.0. > We periodically dump and restore our database. I don't know if an error > occured during a dump restore but our DB seems to be corrupted. > Our base contains many pair of table with the same name. My goodness, that's a pretty obsolete system :-(. You do realize that 7.3.x was up to 7.3.21 by the time it was obsoleted? My recollection is that this has been seen to happen when there were for some reason duplicate rows in pg_shadow --- pg_dump does something like "SELECT FROM pg_class JOIN pg_shadow ON relowner = usesysid", so it would get multiple copies of any table that joined to multiple pg_shadow entries. How you got into that state is harder to say, but if you can get rid of the duplicate entries then your dump would go back to a sane state. regards, tom lane
On 02/08/11 10:20 AM, Tom Lane wrote: > Cyril DE MURCIA<cyril.de-murcia@dcnsgroup.com> writes: >> We use postgresql 7.3.2 under Linux Red Hat 9.0. >> We periodically dump and restore our database. I don't know if an error >> occured during a dump restore but our DB seems to be corrupted. >> Our base contains many pair of table with the same name. > My goodness, that's a pretty obsolete system :-(. You do realize that > 7.3.x was up to 7.3.21 by the time it was obsoleted? indeed. 7.3.2 was released in Feb 2003, 8 years ago, while 7.3.21 was 5 years later, Jan 2008.