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

Re: strange disk access



On Tue, Jan 31, 2006 at 09:07:24PM -0400, BTP wrote:
> I checked my logs and noticed nothing out of the ordinary routine anywwhere.
> At the time I did as you recommend to run top and ps but my system was
> operating so poorly all I got to see was a sad load average of 3+ during all
> the chaos in GNOME...
> 
> Thanks a lot for recommending atop, it looks like a really good tool.
> Although I still need to install the 'cnt' patch to my 2.4 kernel so I can
> track disk and network activity on a per process basis with it. I do not
> have any experience doing this so it should be fun learning. Any quick
> pointers for good tutorials or howtos on debian kernel upgrading etc would
> be appreciated.
> 
> Thanks for the replies
> 
> Bart

Hi Bart,

for a current 2.6 kernel get the sources, the module-init-tools plus 
'kernel-package' like

apt-get install kernel-source-2.6.x module-init-tools kernel-package

('apt-cache search kernel-source' will show you which sources you 
have available at the moment). Then you can read up about compiling 
the kernel the Debian way in /usr/share/doc/kernel-package/README.gz.

Good luck and have fun

Andreas
> 
> On 1/31/06, Andreas Rippl <a.rippl@gmx.net> wrote:
> >
> > On Mon, Jan 30, 2006 at 05:22:55PM -0400, BTP wrote:
> > > Hello, I hope maybe one of you can point me in the right direction to do
> > my
> > > research here...
> > >
> > > on my debian system (on my laptop) for the first time and for no reason
> > my
> > > disk accesses started going crazy and really bogging down the system for
> > > about half a minute. I couldn't even open a terminal window to check the
> > > process tasks...
> > >
> > > How would I go about tracking down exactly what this disk access was all
> > > about? It might have been just swapping but I don't know if it would get
> > > that bad and whether it normally swaps without such problems...
> > >
> > > Bart
> >
> > Hi Bart,
> >
> > my first guess is a cron job. So I'd look in the /etc/cron.* directories
> > (if I knew the time of the disk access). Also in 1.5 minutes, there is
> > plenty of time to start 'top' or 'ps' to see the processes acting up.
> > And finally there is 'atop', which records the running processes. CPU
> > usage etc for later perusal. So you can just let it run and examine what
> > happened later.
> >
> > Hth
> > --
> > Andreas Rippl -- GPG messages preferred
> >                  Key-ID: 0x81073379
> >
> >
> > -----BEGIN PGP SIGNATURE-----
> > Version: GnuPG v1.4.1 (GNU/Linux)
> >
> > iD8DBQFD37KLRs5I5IEHM3kRArQxAKCKxA83QqvFTMbNtsxAtcpd2Ht7lACfU5KY
> > BiKp6iCxZZLdL405xDI8uac=
> > =6T2j
> > -----END PGP SIGNATURE-----
> >
> >
> >

-- 
Andreas Rippl -- GPG messages preferred
                 Key-ID: 0x81073379



Reply to: