Connx Exception error code to catch (issue on Open VMS)

Please take into consideration the exchange here below in order to improve the error code of connx.

De: McGhaw, Larry [mailto:Larry.McGhaw@softwareag.com]
Envoyé: lundi 5 novembre 2018 23:16
À: Jean-Philippe Lutz; MOREAU Stéphane DTSI/DSI
Objet: RE: Connx Exception list

Hello Jean-Philippe,

Thank you for this additional detail, as now I understand the original request.

We do not currently have a list of catchable unique exception codes from Java, as all the exceptions are SQLException or derived from it, and the text contains the error detail.

This is something that could be done, but it is not a simple task, but if it is required I would recommend creating a brainstorm request and this is something we can plan for a future release.

Thanks

lm

From: Jean-Philippe Lutz [mailto:jean-philippe.lutz@ericsson.com]
Sent: Monday, November 5, 2018 10:58 AM
To: McGhaw, Larry <Larry.McGhaw@softwareag.com>; MOREAU Stéphane DTSI/DSI <stephane.moreau@orange.com>
Subject: RE: Connx Exception list

Hi Larry,

If you could provide the list of ConnX specific one this would be a first step

In Stephane example also, the last one is causing a problem

16:07:03 [SELECT - 0 rows, 0.027 secs] [Code: 0, SQL State: HY000] Unable to open KSCL_FTCOBM. Unable to allocate a new socket channel. Maximum number of socket channels per connection reached (1024).

It is too generic. Ideally the best would be to return a dedicated exception like ConnxRuntimeTooManySocketsException. In that way we can cleanly catch the exception and manage a flush to have the adv application resilient for this case. If not possible, at least return a dedicated Code, as it is catch as an SQLException and we can retrieve the errorcode easily.

--- jpl

From: McGhaw, Larry <Larry.McGhaw@softwareag.com>
Sent: mercredi 31 octobre 2018 16:52
To: MOREAU Stéphane DTSI/DSI <stephane.moreau@orange.com>
Subject: RE: Connx Exception list

Hi Stephane,

We do not have an exhaustive list we can send, because in most cases if there is an RMS error, we simply perform a VMS system call and return the RMS error string returned by the operating system:

So in your examples below:

The parts in yellow came from the VMS system call based on the highlighted RMS error code:

15:54:11 [SELECT - 0 rows, 0.053 secs] [Code: 98962, SQL State: HY000] Unable to open KSCL_COANNU. Unable to open table.

>>> STAT:2 ERR:98962 SQLSTATE:S1000 Error opening RMS File. (KSCL_COANNU) %E, file not found

15:54:11 [SELECT - 0 rows, 0.042 secs] [Code: 114690, SQL State: HY000] Unable to open KSCL_FT23PZ. Unable to open table.

>>> STAT:2 ERR:114690 SQLSTATE:S1000 Error opening RMS File. (KSCL_FT23PZ) %E, ACP file access failed

We could create a list based on the VMS documentation – but not 100% sure that is what you want?

Thanks

lm

From: stephane.moreau@orange.com [mailto:stephane.moreau@orange.com]
Sent: Wednesday, October 31, 2018 8:41 AM
To: McGhaw, Larry <Larry.McGhaw@softwareag.com>
Subject: Connx Exception list

Hi Larry,

Could you please provide us all the exception code we may have (on linux side or VMS side) when an issue occurs ? (like quota limit on VMS, file not found, quota limit on linux, other kind of exceptions…)

Here below some example we already encountered :

15:54:11 [SELECT - 0 rows, 0.053 secs] [Code: 98962, SQL State: HY000] Unable to open KSCL_COANNU. Unable to open table.

>>> STAT:2 ERR:98962 SQLSTATE:S1000 Error opening RMS File. (KSCL_COANNU) %E, file not found

15:54:11 [SELECT - 0 rows, 0.042 secs] [Code: 114690, SQL State: HY000] Unable to open KSCL_FT23PZ. Unable to open table.

>>> STAT:2 ERR:114690 SQLSTATE:S1000 Error opening RMS File. (KSCL_FT23PZ) %E, ACP file access failed

16:07:03 [SELECT - 0 rows, 0.027 secs] [Code: 0, SQL State: HY000] Unable to open KSCL_FTCOBM. Unable to allocate a new socket channel. Maximum number of socket channels per connection reached (1024).

Thank you in advance for your reply

Stéphane