Re: V3 protocol + DECLARE problems - Mailing list pgsql-jdbc

From Tom Lane
Subject Re: V3 protocol + DECLARE problems
Date
Msg-id 9141.1090424549@sss.pgh.pa.us
Whole thread Raw
In response to V3 protocol + DECLARE problems  (Oliver Jowett <oliver@opencloud.com>)
Responses Re: V3 protocol + DECLARE problems
List pgsql-jdbc
Oliver Jowett <oliver@opencloud.com> writes:
> Logs follow; basically this is issuing a Parse/Bind/Execute for a
> parameterized DECLARE, which blows up with "no value found for parameter
> 1" despite there definitely being one there. (also, that error appears
> on Execute, not Parse/Bind).

This may be fixable --- I'm thinking that the problem has to do with not
passing down parameters from one portal to another in a situation where
they probably should be.

> Any chance of getting this fixed for 7.5?

I'm up to my keister in PITR and nested-xact issues and won't have time
to look at it soon.  Do you want to take a look and see if you can find
where the ball is getting dropped?

> Alternatively, if we can get WITH HOLD / SCROLL behaviour in portals
> created by Execute (probably means a protocol change) that works too.

This won't be salable unless we decide we really need a protocol change
to support nested xacts properly.  With the change to SAVEPOINT syntax
the motivation for that may have diminished --- in particular I doubt
that we can usefully report a nesting depth, so the urge to fool with
the Z message has diminished greatly (in my mind anyway).

I'd suggest looking into the parameter issue if you want to have
confidence in a fix being available in 7.5.

            regards, tom lane

pgsql-jdbc by date:

Previous
From: "malleshkarthik"
Date:
Subject: unsubscribe
Next
From: Kris Jurka
Date:
Subject: Re: patch: ensure testdbencoding test table is dropped