Pavan Deolasee wrote:
> On 3/27/07, Tom Lane <[EMAIL PROTECTED]> wrote:
>>
>> "Matthijs Melissen" <[EMAIL PROTECTED]> writes:
>> > I am executing the following queries (id has a unique key):
>> > 1) begin;
>> > 1) delete from forum where id = 20;
>> > 1) insert into forum (id, name) values (20, 'test');
>> > 2) delete from forum where id = 20;
>> > 1) commit;
>>
>> > The problem is that process 2 gets the message 'DELETE 0'. I would
>> expect
>> > him to get the message 'DELETE 1'.
>>
>> Why do you find that strange? Process 1 hasn't committed its insert yet.
>>
>>
> I think what he is saying that *after* txn 1 commits, txn 2 does not see
> the
> record inserted by txn1. Isn't that a fair point ? I mean txn 2 can see the
> DELETE operation of txn 1, but can not see the INSERT operation of the
> same transaction.
Not necessarily so. I'd expect only to see a successful delete in txn 2
if a record with id = 20 existed prior to the beginning of txn 1. If
that wasn't the case, then there's nothing to delete in txn 2.
--
Alban Hertroys
[EMAIL PROTECTED]
magproductions b.v.
T: ++31(0)534346874
F: ++31(0)534346876
M:
I: www.magproductions.nl
A: Postbus 416
7500 AK Enschede
// Integrate Your World //
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
