Re: postgres_fdw test timeouts - Mailing list pgsql-hackers

From Nathan Bossart
Subject Re: postgres_fdw test timeouts
Date
Msg-id 20240309142440.GA1111406@nathanxps13
Whole thread Raw
In response to Re: postgres_fdw test timeouts  (Alexander Lakhin <exclusion@gmail.com>)
List pgsql-hackers
On Sat, Mar 09, 2024 at 10:00:00AM +0300, Alexander Lakhin wrote:
> I have re-run the tests and found out that the issue was fixed by
> d3c5f37dd. It changed the inner of the loop "while (PQisBusy(conn))",
> formerly contained in pgfdw_get_result() as follows:
>                 /* Data available in socket? */
>                 if (wc & WL_SOCKET_READABLE)
>                 {
>                     if (!PQconsumeInput(conn))
>                         pgfdw_report_error(ERROR, NULL, conn, false, query);
>                 }
> ->
>         /* Consume whatever data is available from the socket */
>         if (PQconsumeInput(conn) == 0)
>         {
>             /* trouble; expect PQgetResult() to return NULL */
>             break;
>         }
> 
> That is, the unconditional "if PQconsumeInput() ..." eliminates the test
> timeout.

Thanks for confirming!  I'm assuming this just masks the underlying
issue...

-- 
Nathan Bossart
Amazon Web Services: https://aws.amazon.com



pgsql-hackers by date:

Previous
From: Alena Rybakina
Date:
Subject: Re: Support "Right Semi Join" plan shapes
Next
From: Magnus Hagander
Date:
Subject: Re: May be BUG. Periodic burst growth of the checkpoint_req counter on replica.