Re: SPI_ERROR_CONNECT

Thanks Tom,

I sent you a test case.  The problem has since been resolved by changing
one of my functions to VOLATILE instead of IMMUTABLE.  This has caught
me twice now in the last few days.  I hope my learning of this will be a
little more IMMUTABLE :)

cheers,
willem
PG 8.3

Tom Lane wrote:
> Willem Buitendyk <willem@pcfish.ca> writes:
>
>> The problem was with the following:
>>            FOR current_row IN SELECT * from temp_tags_18_counted
>>
>
>
>> The select from the [temp_tags_18_counted]  view is made up of 3 cross
>> joins.  When I simplify and remove the joins everything works.  I tried
>> this with some test data with only a few rows and the joins in place and
>> it works too.
>> In the production data table there are about 250K rows.  Is it possible
>> that calls to queries are colliding here or not giving each other enough
>> time before being whisked around to next call in the FOR loop?
>>
>
> No.  Please provide a test case instead of speculating.  And, again,
> what is the PG version?
>
>             regards, tom lane
>
> ---------------------------(end of broadcast)---------------------------
> TIP 2: Don't 'kill -9' the postmaster
>
>