poor performance hard disk IDE

Stefano Doni stefano.doni at gmail.com
Tue Oct 28 07:53:13 UTC 2008


Hi Charlie,
      yes as I wrote in the bug report, it is against Intrepid:

Hi,
       I've just upgraded to Intrepid and I've seen that a major performance
slowdown is still existing.

It is importat to note that I'm using a 7 years-old PC, so my disks are IDE
(I expect that newer SATA disks don't exibit this behaviour).

I first noticed this performance regression at the time of Gutsy->Hardy
upgrade, and after some search, it appeared to be a common reported problem
probably due to libata and the kernel. Several months ago I found a similar
discussion on LKML, if you can't find it let me know, I'm willing to help to
track down this issue.

Now, the slow disk throughput still exists after the upgrade to Intrepid.
I'm not using any ISO, I've upgraded from Hardy.

As you can see from the bug report, the problem is likely to be due to the
fact that kernel limit my disk to UDMA/33 mode (I've attached a dmesg
output). This seems to be due to the fact that the libata thinks my IDE
cables have 40-wires and not 80, but as I stated in the bug report, this is
incorrect since:

1) Gutsy was having double transfer rate performance with the same hardware
2) I've opened my PC and counted myself the cables :)

That's it! I'll take a look at the debugging procedures, in the meanwhile if
you are willing to help, just ask any info you might need, I'm glad to help
you to track down this issue.

Let me know!

2008/10/27 Charlie Kravetz <cjk at teamcharliesangels.com>

> On Mon, 27 Oct 2008 17:43:37 +0100
> "Stefano Doni" <stefano.doni at gmail.com> wrote:
>
> > Hi,
> >  I can confirm this.
> >
> > Check out this bug I opened regarding Intrepid, but unfortunately it
> > seems that no one has looked at it yet:
> >
> > https://bugs.launchpad.net/ubuntu/+source/linux/+bug/285595
> >
> > Problem exists since Hardy, Gutsy IO bandwidth performance was
> > *double* (measured, not feels).
> >
> > 2008/10/18 shirish <shirishag75 at gmail.com>
> >
> > > Hi all,
> > >  There is a bug regardig poor performance hard disk IDE which has
> > > been filed a year back
> > >
> https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/96693
> > > . The bug seems to be still relevant as my experience shows
> > >
> > >
> > >
> https://bugs.launchpad.net/ubuntu/+source/linux-source-2.6.20/+bug/96693/comments/42
> > >
> > > Can somebody see any issue with my reasoning or anything please
> > > lemme know. --
> > >          Regards,
> > >          Shirish Agarwal
> > >  This email is licensed under
> > > http://creativecommons.org/licenses/by-nc/3.0/
> > >
> > > 065C 6D79 A68C E7EA 52B3  8D70 950D 53FB 729A 8B17
>
> When I read this bug, bug #285595, it appears to be filed against
> Hardy? Please give more information in the bug report, such as what
> version of Ubuntu, if it is intrepid, which ISO image.
>
> Also, take a look at Debugging Procedures,
> https://wiki.ubuntu.com/DebuggingProcedures
> and see if any attachments are needed. The more information given, the
> better chance the bug gets worked.
>
> --
> Charlie Kravetz                         [http://teamcharliesangels.com]
> Linux Registered User Number 425914     [http://counter.li.org/]
> Never let anyone steal your DREAM.      [http://keepingdreams.com]
> bicyclist,  computer tech,  plumber,  customer service rep,  retired,
> diagnosed in 2000 w/ Multiple Sclerosis, MS doesn´t have me down yet...
>
> --
> Ubuntu-devel-discuss mailing list
> Ubuntu-devel-discuss at lists.ubuntu.com
> Modify settings or unsubscribe at:
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ubuntu.com/archives/ubuntu-devel-discuss/attachments/20081028/7d2ca204/attachment.html>


More information about the Ubuntu-devel-discuss mailing list