Re: PATCH/s: RM#1387 - Bad handling of missing connection database server

Hi Dave/Team,

I have attached three patches targeting specific area of work.

-> backend_connection_status_management

* Takes care of the connection status in the psycopg2 driver. And, when the connection is lost, it throws a exception with 503 http status message, and connection lost information in it.

* Contains a change to allow to the flask application to propagate the exceptions even in the release mode.

* Modification for utilising the existing password (while reconnection, if not disconnected explicitly).

* Introduced a new ajax response message 'service_unavailable' (http status code: 503), which suggests temporary service unavailable.

-> manage_client_connect_lost

* It handles the connection lost for different operations at client (browser) side.

* Using the events to handle this situation.

-> remove_connection_status_check

* We're currently return precondition_required (HTTP status code: 428) for the same, but - there is no consistent message/information available to identify it reliably as the server connection lost error (client side changes takes care of this situation, but - this patch is recommended for the whole change to work properly). This patch removes the connection status check from the individual nodes.

TODO:

* Browser tree - node expansion

- When there is an error occurred during expanding the aci-tree node, it does not let the client handle the issue based on the status code, and error message, as it does not expose/share the error details with us. We will have to hack the aci-tree for the same.

* Query Editor

- Slick grid changes are in progress, hence - I did not touch the code for it, as it will lead to rework for either one of us (me, or Murtuza). I will take care of it as soon as the slick grid changes will be committed.

--

Thanks & Regards,

Ashesh Vashi
EnterpriseDB INDIA: Enterprise PostgreSQL Company

http://www.linkedin.com/in/asheshvashi