Re: [HACKERS] MERGE SQL Statement for PG11

On 29 January 2018 at 17:35, Peter Geoghegan <pg(at)bowt(dot)ie> wrote:
> On Mon, Jan 29, 2018 at 8:51 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
>> On 29 January 2018 at 16:44, Bruce Momjian <bruce(at)momjian(dot)us> wrote:
>>
>>> I think the question is how does it handle cases it doesn't support?
>>> Does it give wrong answers? Does it give a helpful error message? Can
>>> you summarize that?
>>
>> I'm happy to report that it gives correct answers to every known MERGE
>> test, except
>>
>> * where it hits a concurrency issue and throws SQLCODE =
>> ERRCODE_T_R_SERIALIZATION_FAILURE and the standard text for that
>>
>> * where it hits an unsupported feature and throws SQLCODE =
>> ERRCODE_FEATURE_NOT_SUPPORTED, with appropriate text
>
> What specific features does it not work with already? A list would be helpful.

Yes, I added that to the docs as a result of your review comments.

I also mentioned them here last week in your review in answer to your
specific questions.

The current list of features that return ERRCODE_FEATURE_NOT_SUPPORTED is
* Tables with Row Security enabled
* Partitioning & Inheritance
* Foreign Tables

Rules are ignored, as they are with COPY.

If people have concerns or find problems following review, I will be
happy to update this list and/or fix issues, as normal.

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