On 03/30/18 07:39, Laurenz Albe wrote:
>> psql:testing/test.pg_sql:41: NOTICE: Connection busy: 1
>>
>> psql:testing/test.pg_sql:41: NOTICE: Last error: OK
>>
>> psql:testing/test.pg_sql:41: NOTICE: Cancel query: OK
>>
>> psql:testing/test.pg_sql:41: NOTICE: Connection busy: 0
>>
>> psql:testing/test.pg_sql:41: NOTICE: could not send query: another
>> command is already in progress
>
>> Has anyone an idea?
>
> The cause of the error message is clear; as the documentation says:
Not to me. As mentioned in my first post, originally I did no cancelling
and error checking and had the same error. Trying to cancelling did not
alter behaviour. It boils down to: Why do I get the "another command is
already in progress" when dblink_is_busy acknowledges the dblink to not
be busy by returning 0?
> the query will fail soon. You must still complete the normal query protocol,
> for example by calling dblink_get_result.
Ah, maybe this is the underlying problem. If dblink requires that
results get retrieved by dblink_get_result before the dblink is actually
ready to receive another query, it would explain the error I get.
However, I feel than the result of dblink_is_busy is faulty,
counter-intuitive or just useless in that context. Or I just
misinterpreted documentation: "checks if connection is busy with an
async query"
My understand there is that the actual query is still being processed,
the gathering of the results. I did not count the keeping of the result
as part of the query.
I hope I can check on that today and get back with my findings.
--
SIP/iptel.org: thiemo.kellner
Öffentlicher PGP-Schlüssel:
http://pgp.mit.edu/pks/lookup?op=get&search=0xCA167FB0E717AFFC