PostgreSQL 9.3 benchmarks

Raimundo Santos raitech at gmail.com
Mon Mar 10 20:39:42 PDT 2014


Hello all!

Just to note: I have seen an annoying bug with FreeBSD in Gigabyte and
Intel motherboards with Intel processor Core i5 and i3 and USB 3.0.

When I remove the monitor from D-Sub or DVI connectors, XHCI drivers got to
hundreds of thousands of interrupts/sec and stays there. Need reboot to
stop it. Like this: http://www.freebsd.org/cgi/query-pr.cgi?pr=179376.
Could reproduce in FreeBSD 10.0 too.

But my hardware, what can I say, is cheap hardware.

Fracois, have you removed a monitor from the test machine (this is really
strange to ask, I know)? Motherboard has USB 3?

May this behaviour interfere with VM faults?

Best regards,
Raimundo Santos



On 10 March 2014 05:29, Francois Tigeot <ftigeot at wolfpond.org> wrote:

> Hi,
>
> I took the opportunity of having access to a 20 cores / 40threads Xeon
> system
> to do some Postgres benchmarking.
>
> The goal was to test DragonFly scalability with a read-only workload; I
> also
> threw in some operating systems often used in server environments for
> comparison purposes.
>
> I must say I'm quite pleased with DragonFly's position in the results :-)
>
> --
> Francois Tigeot
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dragonflybsd.org/pipermail/users/attachments/20140311/9c13d2ff/attachment-0013.html>


More information about the Users mailing list