Re: [HACKERS] PgFDW connection invalidation by ALTER SERVER/ALTERUSER MAPPING - Mailing list pgsql-hackers

From Michael Paquier
Subject Re: [HACKERS] PgFDW connection invalidation by ALTER SERVER/ALTERUSER MAPPING
Date
Msg-id CAB7nPqQ_a5R7M5baLEA3cOwCAs5AjxVO62HUdpJJvO7oeB2umQ@mail.gmail.com
Whole thread Raw
In response to Re: [HACKERS] PgFDW connection invalidation by ALTER SERVER/ALTERUSER MAPPING  (Alvaro Herrera <alvherre@2ndquadrant.com>)
List pgsql-hackers
On Fri, Jul 21, 2017 at 12:23 AM, Alvaro Herrera
<alvherre@2ndquadrant.com> wrote:
> Kyotaro HORIGUCHI wrote:
>> Finally, I added a new TAP test library PsqlSession. It offers
>> interactive psql sessions. Then added a simple test to
>> postgres_fdw using it.
>
> Hmm, I think this can be very useful for other things.  Let's keep this
> in mind to use in the future, even if we find another way to fix the
> issue at hand.  In fact, I had a problem a couple of weeks ago in which
> I needed two concurrent sessions and one of them disconnected in the
> middle of the test.

Agreed, I wanted the ability to hold a session at hand a couple of
times already for tests. And I agree with the point of having a
separate discussion for such things out of the scope of a bug fix.
Thinking larger, I think that it would be more helpful to hold
processes and run commands in parallel, say for pg_receivewal.

> Can't do that with isolationtester ...

In the pglogical fork of Postgres, you guys improved isolationtester
to handle multiple hosts, right? That sounds harder to integrate than
a perl module though, as isolation tester starts only one server.
-- 
Michael



pgsql-hackers by date:

Previous
From: Kyotaro HORIGUCHI
Date:
Subject: Re: [HACKERS] PgFDW connection invalidation by ALTER SERVER/ALTERUSER MAPPING
Next
From: Neha Sharma
Date:
Subject: Re: [HACKERS] [TRAP: FailedAssertion] causing server to crash