[HACKERS] Trigger problem - conclusion

[Nikolay Grebnev]

Good Day,

I recently posted a message here (
http://archives.postgresql.org/pgsql-hackers/2007-12/msg00340.php) that the
trigger does not work as it should from time to time. Now the trigger works
on C, before It was on TCL and it had the same problem.
As the trigger works all right in 99.999 % cases (or even more) then we can
make a conclusion that the trigger is written all right. And it was written
in complete accordance with the documentation that is possible to find.
I suggest to write down in documentation for PostgreSQL, that during big
loads triggers can fail in some cases.

Nik

Re: [HACKERS] Trigger problem - conclusion

[Andrew Dunstan]

Nikolay Grebnev wrote:

Good Day,

I recently posted a message here 
(http://archives.postgresql.org/pgsql-hackers/2007-12/msg00340.php) 
that the trigger does not work as it should from time to time. Now the 
trigger works on C, before It was on TCL and it had the same problem.
As the trigger works all right in 99.999 % cases (or even more) then 
we can make a conclusion that the trigger is written all right. And it 
was written in complete accordance with the documentation that is 
possible to find.
I suggest to write down in documentation for PostgreSQL, that during 
big loads triggers can fail in some cases.





I think you need to provide a reproducable, self-contained test case, if 
possible.


Your previous mail said:

In php where the queries are formed we see that the query is sent for 
execution and executed ant the base was modified, but the trigger 
seems just to skip it (does not work with the query). 


It would be far better to check that the statement has executed in the 
log, after turning on log_statement or log_min_duration_statement.


Does it fail in inserts, updates, deletes, or all three?

What platform are you using?

And why are you compiling with -O3?

cheers

andrew


---(end of broadcast)---
TIP 3: Have you checked our extensive FAQ?

  http://www.postgresql.org/docs/faq

[HACKERS] Trigger problem

[jreniz]

Hi friends!

I'm working in a trigger and I need to put the result
of a query into a variable.

That's very easy- apparently!

The query has a aggregate function like this:

select sum(field) into variable ...

and I'm sure that field and variable are int4 type.

So, when I run this trigger there is a mistake:
 ''there is no operator '=$' for types 'int4' and 'int4'
   you will either have to retype this query using an
   explicit cast, or you will have to define the operator
   using CREATE OPERATOR''

what's meaning this? and
how can I assign the result of aggregate function into a
variable?
(My system is 6.5.3)




---(end of broadcast)---
TIP 6: Have you searched our list archives?

http://www.postgresql.org/search.mpl

Re: [HACKERS] Trigger problem

[Tom Lane]

jreniz [EMAIL PROTECTED] writes:
 So, when I run this trigger there is a mistake:
  ''there is no operator '=$' for types 'int4' and 'int4'

 (My system is 6.5.3)

This is an old bug.  Update to 7.0.3.

It might work to add spaces around the '=' signs in your trigger
function, but an update would be a good idea anyway.

regards, tom lane

---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])