Re: Improving replay of XLOG_BTREE_VACUUM records

On 10 March 2016 at 06:27, Michael Paquier <michael(dot)paquier(at)gmail(dot)com>
wrote:

> On Thu, Mar 10, 2016 at 1:29 AM, David Steele <david(at)pgmasters(dot)net> wrote:
> > On 1/8/16 9:34 AM, Alvaro Herrera wrote:
> >> Simon Riggs wrote:
> >>>
> >>> On 8 January 2016 at 13:36, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>
> >>> wrote:
> >>>>
> >>>> I would agree except for the observation on toast indexes. I think
> >>>> that's an important enough use case that perhaps we should have both.
> >>>
> >>> The exclusion of toast indexes is something we can remove also, I have
> >>> recently discovered. When we access toast data we ignore MVCC, but we
> >>> still
> >>> have the toast pointer and chunkid to use for rechecking our scan
> >>> results.
> >>> So a later patch will add some rechecks.
> >>
> >> Ah, interesting, glad to hear. I take it you're pushing your patch
> >> soon, then?
> >
> > ISTM that this patch should be "returned with feedback" or "rejected"
> based
> > on the thread. I'm marking it "waiting for author" for the time being.
>
> I think that we are still waiting for some input from Simon here...
> Simon, are you going to finish wrapping up your other patch?
>

Yes, I have done the research, so think patch should be rejected now.

Thanks to everyone for their input. It's good to have alternate approaches.

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