Re: [HACKERS] Re: Clarifying "server starting" messaging in pg_ctlstart without --wait

* Michael Paquier (michael.paquier@gmail.com) wrote:
> On Fri, Jan 20, 2017 at 12:06 AM, Stephen Frost <sfrost@snowman.net> wrote:
> > We did make the WAL checksum routines a lot
> > faster with 9.6, as I recall, so perhaps there's been some change there
> > too.
>
> 9.5, commit 5028f22f with Abhijit's and Heikki's work on CRC-32 computations.

Oh, was it 9.5 where we added the code to use CPU OPs to speed up the
calculations..?  If so then that surprises me a bit, as, if I recall
correctly, I had a 9.5-based server where the replica simply couldn't
keep up with replay and was entirely CPU-bound..

I had thought those changes went into 9.6.  If they were in 9.5 then
that's possibly a bit disappointing.

Of course, this was a while ago, perhaps that was actually a 9.4
server..  mmm, will have to go find out.

Thanks!

Stephen