Closed tsegismont closed 2 months ago
no test ?
@vietj I searched the web for cases when the server would raise a notice/error not related to a client request, but I couldn't find any. Have you encountered this case yourself before?
I think it can be reproduced using a proxy, we have similar tests like that
you can look at testDisconnectAbruptlyDuringStartup
I think you misunderstood my comment, the problem is not to test re-connection, the problem is to force the PostgreSQL server to raise a notice message while there are no inflight commands on the connection.
Afaiu this is not the case in testDisconnectAbruptlyDuringStartup
yes, Im saying to use the proxy server to create a fake message to be sent to the client using it
On Thu, Aug 29, 2024 at 3:19 PM Thomas Segismont @.***> wrote:
I think you misunderstood my comment, the problem is not to test re-connection, the problem is to force the PostgreSQL server to raise a notice message while there are no inflight commands on the connection.
Afaiu this is not the case in testDisconnectAbruptlyDuringStartup
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you were mentioned.Message ID: @.***>
@vietj PTAL
Thanks for the tip
See #1442
This may happen if the connection is used by a PgSubscriber.