Re: BUG #16145: Not able to terminate active session - Mailing list pgsql-bugs

From Tomas Vondra
Subject Re: BUG #16145: Not able to terminate active session
Date
Msg-id 20191204135536.3drofgso4sut5enj@development
Whole thread Raw
In response to Re: BUG #16145: Not able to terminate active session  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: BUG #16145: Not able to terminate active session
List pgsql-bugs
On Wed, Dec 04, 2019 at 08:34:34AM -0500, Tom Lane wrote:
>MOHAN KUMAR DORAIRAJ <mohankumar.dorairaj@globalfoundries.com> writes:
>> Is there any other alternative option available.?
>
>Well, why is it that your session is sitting waiting for input from
>the Oracle server?  That doesn't seem like a condition that ought
>to be persistent, most of the time.  Perhaps redesigning your
>queries against that foreign server would help.
>

Yeah, having some CHECK_FOR_INTERRUPT calls in the code might help
(there's a single place with this macro, but it only happens after an
error). The question is whether the OCIStmtExecute actually returns from
time to time, of it if gets stuck.

>Also, I do recommend asking the Oracle FDW authors about this.
>It's possible that adding some CHECK_FOR_INTERRUPT calls in that
>FDW would ameliorate things.  (I don't know where to file issues
>about oracle_fdw, but it's not this list.)
>

I think the right place to file the issue is here:

   https://github.com/laurenz/oracle_fdw

regards

-- 
Tomas Vondra                  http://www.2ndQuadrant.com
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services



pgsql-bugs by date:

Previous
From: Alexey Bashtanov
Date:
Subject: Re: pg_dump losing index column collations for unique and primarykeys
Next
From: Tom Lane
Date:
Subject: Re: BUG #16147: postgresql 12.1 (from homebrew) - pg_restore -h localhost --jobs=2 crashes