Re: Fwd: Problem with a "complex" upsert

Andres Freund <andres@anarazel.de>

From: Andres Freund <andres@anarazel.de>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Dean Rasheed <dean.a.rasheed@gmail.com>, Amit Langote <Langote_Amit_f8@lab.ntt.co.jp>, Mario De Frutos Dieguez <mariodefrutos@gmail.com>, pgsql-bugs@lists.postgresql.org
Date: 2018-08-03T22:44:55Z
Lists: pgsql-bugs
On 2018-08-03 18:32:57 -0400, Tom Lane wrote:
> Dean Rasheed <dean.a.rasheed@gmail.com> writes:
> > On 3 August 2018 at 19:46, Andres Freund <andres@anarazel.de> wrote:
> >> Are you planning to push a version of this soon? It'd be good to
> >> get this included in the next set of releases...
> 
> > Yes, agreed. I'll try to do it this weekend.
> 
> Keep in mind that we are hard up against the release deadline.

Right.


> This is a bad weekend to be pushing anything you are not 100.00%
> sure of; the later, the more so, as by Sunday you will probably not
> get a complete set of buildfarm reports before the wrap happens.
> 
> Balance the risks of shipping a broken release vs. waiting one
> more quarter to ship the fix.  After a quick look at the size
> of the patch, my own inclination if I were the committer would
> be to wait till after the releases are out.  Or you might
> consider pushing only to 11+HEAD, with the expectation of
> back-patching later after we've gotten some beta results.

This results in clearly inserting wrong data and/or crashing the
server. And it's not a huge effect outside of already broken scenarios.
I think we definitely should try to get this in.

Greetings,

Andres Freund


Commits

  1. Fix INSERT ON CONFLICT UPDATE through a view that isn't just SELECT *.