Hi Tom,
Ehr...
The problem must have been the heat...
I forgot to set the sequences to the right values after inserting the data.
This is something that is done automatically in our testprocedure, but in
this windows test i forgot.
Did it the right way now, the errors are history.
-- sigh --
BTW: We (and the sales guys) are absolutely delighted with the arrival of
the native windows version!
Greetings, Joek
=20
-----Oorspronkelijk bericht-----
Van: Tom Lane [mailto:tgl@sss.pgh.pa.us]=20
Verzonden: zaterdag 14 augustus 2004 2:13
Aan: J. Hondius
CC: pgsql-bugs@postgresql.org
Onderwerp: Re: [BUGS] Sequence problem in stresstest on 8.0-beta1 (windows)
"J. Hondius" <jhondius@rem.nl> writes:
> These statements:
> insert into "tbverantwoording" ("dnkeyaanstelling", "dnjaarweek",
> "dnkeyactiviteitenschema") values (27, 200343, 1781); insert into=20
> "tbverantwoording" ("dnkeyaanstelling", "dnjaarweek",
> "dnkeyactiviteitenschema") values (27, 200343, 1780);
> Reported these errors:
> psql:sqlfiles/0.sql:241: ERROR: duplicate key violates unique=20
> constraint "pk_tbverantw"
> psql:sqlfiles/0.sql:223: ERROR: duplicate key violates unique=20
> constraint "pk_tbverantw"
> BUT ONLY DURING HIGH LOAD, 717 errors where reported, for 2000=20
> inserts.
I couldn't replicate any problem. I put 5000 copies of those two statements
into a file and set up half a dozen psql sessions concurrently reading the
file. No errors.
If you can provide a complete, reproducible test case, I'll be glad to look
at it, but there's not enough information here to do anything ...
regards, tom lane
