Re: Why F15 has up to 50% worse disk performance than Ubuntu ?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, 2011-06-07 at 17:30 +0100, Alan Cox wrote: 
> On Tue, 7 Jun 2011 16:20:32 +0000 (UTC)
> JB <jb.1234abcd@xxxxxxxxx> wrote:

> > Alan Cox <alan <at> lxorguk.ukuu.org.uk> writes:

Alan!  Good lord, it's been years, dude...  Too many years since ALS.
Too many years not making it to LinuxWorld (for me).  Sigh...

> > > ...
> > > Equally if Fedora has soem crappy background app that keeps writing
> > > little bits of pointless data to the disk every second that'll do the job
> > > nicely too.
> > 
> > I think you did not make this comment out of the blue ?

> I've not looked into F15 disk performance, and as I never plan to run
> FC15 I doubt I ever will. I'm skipping this release.

Do you really feel this is wise?  I test every Fedora release
defensively.  I want to know what they are screwing up before it's cast
in stone and I get burned by it down the road.  F15 is certainly
building up a list on my shit-list already.  I treat Ubuntu the same
way.  Test the 6 month on at least one machine after giving it a month
for the dirt and debris to settle.

If you look back through this list you'll find one of my rants on
preupgrade (sucks less? sucks more?)  Preupgrade basically destroyed a
machine under me thanks to (I believe) a dependency conflict with
avidmux that it failed to catch and that "yum distro-sync" caught and
handled gracefully on another machine that survived the upgrade.

They've also screwed up IPv6 autoconf over bridges needed for virtual
machines (mentioned in the aforementioned thread) and I'm about to file
a bug report on that.  (Ok, ok...  Messing with IPv6.  Now THAT really
pisses me off...)

So far, I haven't experienced any disk performance degradation but I'll
start checking (on the machines that survived the upgrade...  Sigh...).
One of them is one of my development engines with a lot of memory, a lot
of virtual machines and a whole lotta load.  It better work and better
work good.

They've done enough in there to give it a careful look-over even if it's
on a machine you don't care about.  I'm really REALLY unhappy with
systemd and the machine that got slagged, ATM.  It's dumping me in
"Emergency Mode" and running "systemctl default" reports it could not
bring up the method "Transaction would be destructive." like what the
hell is THAT???  I'm withholding judgment but feel like systemd has the
makings of the biggest steaming pile since the early days of
NetworkMangler but I'm not reaching for the wooden stake and mallet just
yet.  Since the early days of NetworkMangler, I've even done some work
on the NetworkManager plugins (vpnc and openswan).  Even steaming piles
can fertilize productive end results.  But, then again, I'm old school.
Give me scripts I can edit and debug.  Don't cripple my system with
binaries that fsck-up and don't give me coherent errors.

> I can't specifically say FC15 has that problem but its something I've
> seen chasing down I/O performance problems on systems. It doesn't
> take much to give a disk a pile of extra seeks to do and that hurts
> because seek times haven't changed in years and are now relatively very
> expensive.

I've seen crypto do this or some configurations of raid and MD.  Your
absolutely correct.  I've seen a change in an order of magnitude there
just from minor tweaks and changes.  Long gone are the days we worried
about and tweaked the interleave of drives for the spindle speed.  The
concept and effects remain with us and the impact remains the same.

> > So why not bring it (them) into the light and knock around a bit ?
> > Btw, would that apply to kernel daemons as well ?

> I've no interest in Fedora 15 and pinning it down is a fairly big job
> involving various re-installs.

> > Also it'd be interesting to know if you kept the OS default partitioning scheme
> > (in which case the Fedora install would likely have an LVM layer the Ubuntu one
> > would not.)
> > 
> > Yes, defaults.

> Could be. Still fair benchmark because Fedora configures the disks that
> way default.

Yeah...  Would also be interesting to compare a fresh install to an
upgrade.

> Easy way to check - go do two installs on a box similar to the Phoronix
> one and see what you get if you want to find out.

> Alan

Regards,
Mike
-- 
Michael H. Warfield (AI4NB) | (770) 985-6132 |  mhw@xxxxxxxxxxxx
   /\/\|=mhw=|\/\/          | (678) 463-0932 |  http://www.wittsend.com/mhw/
   NIC whois: MHW9          | An optimist believes we live in the best of all
 PGP Key: 0x674627FF        | possible worlds.  A pessimist is sure of it!

Attachment: signature.asc
Description: This is a digitally signed message part

-- 
users mailing list
users@xxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe or change subscription options:
https://admin.fedoraproject.org/mailman/listinfo/users
Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
[Index of Archives]     [Older Fedora Users]     [Fedora Announce]     [Fedora Package Announce]     [EPEL Announce]     [EPEL Devel]     [Fedora Magazine]     [Fedora Summer Coding]     [Fedora Laptop]     [Fedora Cloud]     [Fedora Advisory Board]     [Fedora Education]     [Fedora Security]     [Fedora Scitech]     [Fedora Robotics]     [Fedora Infrastructure]     [Fedora Websites]     [Anaconda Devel]     [Fedora Devel Java]     [Fedora Desktop]     [Fedora Fonts]     [Fedora Marketing]     [Fedora Management Tools]     [Fedora Mentors]     [Fedora Package Review]     [Fedora R Devel]     [Fedora PHP Devel]     [Kickstart]     [Fedora Music]     [Fedora Packaging]     [Fedora SELinux]     [Fedora Legal]     [Fedora Kernel]     [Fedora OCaml]     [Coolkey]     [Virtualization Tools]     [ET Management Tools]     [Yum Users]     [Yosemite News]     [Gnome Users]     [KDE Users]     [Fedora Art]     [Fedora Docs]     [Fedora Sparc]     [Libvirt Users]     [Fedora ARM]

  Powered by Linux