Duplicate records???

[ptillemans]

Hi,

During an upgrade I came accross 2 (actually 4) records which I cannot distinguish, but which are duplicates.

tvl_test=# select namespace,id,max(oid),min(oid) from tvlprodcounter a group by namespace,id having count(*)>1;
namespace | id | max | min
-------------------------------------+--------------------+--------+--------
com.synegorie.adverts.ejb.AdvertCMP | etri_timesClicked | 238225 | 238225
com.synegorie.adverts.ejb.AdvertCMP | happy_timesClicked | 238364 | 238364
(2 rows)


Does anybody know how I can separate them and delete just 1 of the 2?

Thanks,

Peter

 

[Leland123]

Hi ptillemans,

There is a special column in all postgres tables named OID. This is a number and is usually unique to each row in the table, except in very large tables the OID can wrap around and be reused. The following would display the OID in one of my tables:

select the_key, first, last, OID from testorders;

I would recommend that you have a serial field in every table that is assigns a unique key from an auto incrementing sequence. Look at the postgres doc to learn more about setting up a sequence.

Leland F. Jackson, CPA
Software - Master (TM)

https://www.smvfp.com

Nothing Runs Like the Fox