To whom it may concern:
I found a constraint bug on PostgreSQL
lastest version (7.2).
Here is how I got the
bug.
------------------------------------------------------------------------------------------------ testdb=> testdb=> create table mytest( testdb-> id int primary key, testdb-> masterid int constraint fk_mytest_constr1 REFERENCES mytest(id) ON DELETE CASCADE ON UPDATE NO ACTION DEFERRABLE INITIALLY DEFERRED testdb-> ); NOTICE: CREATE TABLE / PRIMARY KEY will create implicit index 'mytest_pkey' for table 'mytest' NOTICE: CREATE TABLE will create implicit trigger(s) for FOREIGN KEY check(s) CREATE testdb=> begin; BEGIN testdb=> insert into mytest(id,masterid) values(1,null); INSERT 447153 1 testdb=> insert into mytest(id,masterid) values(2,5); INSERT 447154 1 testdb=> insert into mytest(id,masterid) values(3,2); INSERT 447155 1 testdb=> update mytest set masterid=1 where id=2; UPDATE 1 testdb=> end; ERROR: fk_mytest_constr1 referential integrity violation - key referenced from mytest not found in mytest testdb=> ------------------------------------------------------------------------------------------------ Eric
PS. 1.There won't be such an error if I do another insert statement which doesn't violate the reference constraint before ending the transaction.(eg.insert into mytest(id,masterid) values(4,3);) 2.It happens while I was using ver7.1.3 and I do an upgrade and run the regression test(everything is fine). 3.It's on a pentium 75(run as 120MHZ) with 40MB RAM, Redhat 6.1 |
- Re: [BUGS] referential constraint bug Eric Lu
- Re: [BUGS] referential constraint bug Stephan Szabo
- Re: [BUGS] referential constraint bug Bruce Momjian
- Re: [BUGS] referential constraint bug Stephan Szabo
- Re: [BUGS] referential constraint bug Bruce Momjian
- Re: [BUGS] referential constraint bug Bruce Momjian