Re: 15 pg_upgrade with -j

On 2023-05-22 21:10:48 -0500, Ron wrote:
> On 5/22/23 18:42, Tom Lane wrote:
> > It looks like the assumption was that issuing link()
> > requests in parallel wouldn't help much but just swamp your disk
> > if they're all on the same filesystem.
> > Maybe that could use rethinking, not sure.
>
> It does need rethinking in the era of VMs and SANs. /var/lib/pgsql/15 is
> going to be on a different LUN from /var/lib/pgsql/9.6

You can't hardlink between different file systems.

Even if you could assign single directories to specific LUNs (does any
file system allow this?) this would at best spread the updates across
two LUNs (the inodes would presumable stay on the source LUN and the
target directory would be on the target LUN).

hp

--
_ | Peter J. Holzer | Story must make more sense than reality.
|_|_) | |
| | | hjp(at)hjp(dot)at | -- Charles Stross, "Creative writing
__/ | http://www.hjp.at/ | challenge!"