Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > alt.os.linux.slackware > #26650

Re: HD writes hog RAM with SATA HD replacing an IDE

Show all headers | View raw

Rich <rich@example.invalid> writes:

> In alt.os.linux.slackware Mike Spencer <mds@bogus.nodomain.nowhere> wrote:
> 
>> After reinstall of Slackware 11 [1]
> 
> Your issue likely arises here.  Slack 11 is very old, with a very old
> kernel.

Yes. Several others have replied "Old! Old! Gak!".  I know.
No need for me to rant here about *why* I have this old machine on an
old system.  

>> The drive is seen by Linux as hda, not sda. (Some on-mobo emulation I
>> don't know about?)
> 
> Or, more likely: this suggests that the kernel included in slack 11 is
> prior to the retirement of the legacy hd[a-z] driver for IDE disks, and
> the kernel is running your new sata port in legacy IDE mode (because
> that is all it (this specific kernel) knows how to do).

Looking at the source, it appears that the 2.4.33 kernel knows about
SATA.  The oddity is that the IBM mobo has those "SATA IDE n" plugs
and seems to emulate IDE.

> And, if this is the issue, you may even have the slack 11 kernel
> running the sata port in IDE emulation mode using PIO, which would
> produce exactly the issue you describe (large latencys and system
> pauses on disk accesses).

Possibly.  I don't know how to determoine that.  But it's only disk
write that bogs down. Disk read behaves normally.  And PIO would boot
CPU usage, wouldn't it? I don't think I'm seeing that.

> You'd do well to upgrade to a more modern Slackware, and you'd likely
> find that this issue disappears when you do so.

When it's unavoidable.  I have Slack 12 on another machine and 14.1 on
a laptop.  On the 14.1 system, I've spent hours defeating new
"features" that I don't want.

Before I do any more grovelling through kernel parameter and such that
I don't fully understand, tomorrow I'm about to increase the RAM in
the hope that the excessive buffering of writes that appears to be
hogging memory will be self-limiting.

-- 
Mike Spencer                  Nova Scotia, Canada

Back to alt.os.linux.slackware | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

HD writes hog RAM with SATA HD replacing an IDE Mike Spencer <mds@bogus.nodomain.nowhere> - 2016-04-11 02:47 -0300
  Re: HD writes hog RAM with SATA HD replacing an IDE Thomas Richter <thor@math.tu-berlin.de> - 2016-04-11 09:14 +0200
    Re: HD writes hog RAM with SATA HD replacing an IDE root <NoEMail@home.org> - 2016-04-11 16:54 +0000
  Re: HD writes hog RAM with SATA HD replacing an IDE JohnF <john@please.see.sig.for.email.com> - 2016-04-11 07:22 +0000
  Re: HD writes hog RAM with SATA HD replacing an IDE philo <philo@privacy.net> - 2016-04-11 04:19 -0500
    Re: HD writes hog RAM with SATA HD replacing an IDE Mike Spencer <mds@bogus.nodomain.nowhere> - 2016-04-13 00:06 -0300
      Re: HD writes hog RAM with SATA HD replacing an IDE philo <philo@privacy.net> - 2016-04-13 13:26 -0500
  Re: HD writes hog RAM with SATA HD replacing an IDE "Carlos E.R." <robin_listas@invalid.es> - 2016-04-11 11:30 +0200
  Re: HD writes hog RAM with SATA HD replacing an IDE Joe Beanfish <joebeanfish@nospam.duh> - 2016-04-11 13:17 +0000
  Re: HD writes hog RAM with SATA HD replacing an IDE Rich <rich@example.invalid> - 2016-04-11 15:43 +0000
    Re: HD writes hog RAM with SATA HD replacing an IDE Mike Spencer <mds@bogus.nodomain.nowhere> - 2016-04-12 23:42 -0300
      Re: HD writes hog RAM with SATA HD replacing an IDE Rich <rich@example.invalid> - 2016-04-13 03:29 +0000
        Re: HD writes hog RAM with SATA HD replacing an IDE Mike Spencer <mds@bogus.nodomain.nowhere> - 2016-04-15 03:08 -0300

csiph-web