On 13-May-2014 18:19 -0500, Sarah Kemp wrote:
I will keep my question here brief and link to a more detailed
version so the code is readable.
I am implementing SQL error handling based on this page:
GET DIAGNOSTICS is incorrectly returning an SQLSTATE of 00000 when
it is first called in a newly activated service program. All
subsequent calls return correctly. Executing an otherwise unnecessary
SQL statement beforehand eliminates the issue. The service program
containing the GET DIAGNOSTICS statement is bound-by-reference to the
service program containing the erroring SQL statement and activates
*IMMED in group *CALLER.
Please see my full question here:
I will be happy to answer my own question on that site with any
information sent to this list.
The described effect seems suspect, so very possibly is indicative of
a defect. As such, quite probably the best path would be to pursue the
issue with your service provider. The most recent conspicuous fix
activity involving the output from a GET DIAGNOSTICS statement is
available in v7r1 PTF SI51967. Note: nowhere did I see any mention of
the version, release, mod, tech refresh, or cumulative level; thus a
recommendation of a IBM i 7.1 fix may not be germane.