Re: Gnome is very slow

[Chris Rouch <crouch pobox com>]

On Mon, 16 Feb 2004 17:09:43 -0600
"Hoyt Bailey" <hoyt13 wiredok com> wrote:

> 
> ----- Original Message ----- 
> From: "Chris Rouch" <crouch pobox com>
> To: <gnome-list gnome org>
> Sent: Monday, February 16, 2004 11:09
> Subject: Re: Gnome is very slow
> 
> 
> > On Mon, 16 Feb 2004 10:35:37 -0600
> > "Hoyt Bailey" <hoyt13 wiredok com> wrote:
> >
> > >
> > > ----- Original Message ----- 
> > > From: "Hoyt Bailey" <hoyt13 wiredok com>
> > > To: "gnome-list" <gnome-list gnome org>
> > > Sent: Sunday, February 15, 2004 16:34
> > > Subject: Gnome is very slow
> > >
> > > *********UPDATE*******
> > > Apparantly it isnt a network problem,  because I can ping
> > > localhost and get satasfactory response 0 errors @0.26 to 0.29ms
> > > and hostname returns localhost.  Another suggestion was made
> > > earlier to try strace.
> > >  Well I tried
> > > and I dont know how to involke strace 'strace -Tcf /tmp/strace.o'
> > > wasnt adequate niether was 'strace -Tcfo /tmp/strace'.  Therefore
> > > if you would like to see the output I need a workable command. 
> > > The man page indicates that the output can be limited,  this might
> > > be a good idea but 2 minutes is a long time @ 1.7Ghz.
> > >
> >
> > I don't think strace is going to help, but here are some things you
> > can try:
> >
> > 1) ssh localhost uptime
> >
> > This will either prompt you for a password or complain that it can't
> > connect to a sshd. In either case, the response should be very quick
> > - if it takes 30 seconds or so the the problem is with the network.
> >
> 
> Issued 'ssh localhost uptime' as me.  From terminal in KDE & gnome.
> ssh: Connect to localhost port 22.  Connection refused
> Issued 'ssh localhost uptime' as su (root).
> ssh: Connect to localhost port 22.  Connection refused
> In all cases response was instant.  Refusal  could be due to High
> Security setting.
> 
> > 2) df
> > If this hangs then you have nfs/automounter problems. If it finishes
> > quickly then you don't. Check that none of your partitions are close
> > to 100% full
> >
> 
> Issued df
> /dev/ide/host0/bus0/target0/lun0/part3
> Size 65G, Used 6.5G, Avail 59G, Use 10%, Mounted on /
> /dev/ide/host0/bus0/target0/lun0/part1
> Size 47G, Used 11G, Avail 36G, Use 23%, Mounted on /mnt/windows
> 
> > 3) uptime
> > Make sure that the 3 load average numbers are all less than 1. If
> > they're not, wait 5 minutes and do it again. If they're still too
> > big then you probably have some rogue processes.
> >
> 
> Only results are from top.
> 
> > 4) look at .gnomerc-errors and .xsession-errors. Maybe there is
> > somthing nasty showing up there.
> >
> See bottom comments.
> 
> > 5) look at the output of top. Are there processes hogging the cpu?
> > Are you nearly out of memory or swap space?
> >
> 
> top
> Tasks 91 total 1 running 90 sleeping 0 stopped 0 zombie
> cpu 2.3%, User 1.7%, System 0.0%, Nice 0.0%, Idle 97.3%.
> mem 514132K total, 261980K used, 252152K free, 22456 Buffers.
> Swap 530104K, 0K used, 530104K free, 182808 cached
> 
> > Finally, is it just nautilus that is slow or does the whole system
> > seem sluggish? I haven't used nautilus for a long time, but one of
> > the reasons I stopped was because it was too slow. This was several
> > versions ago though (gnome 2.0 I think).
> >
> I could not locate .gnomere-errors. .xsession-errors I intended to
> attach, however I was unable to transfer to either floppy or to a CD. 
> There were two warnings listed both repeated a number of times.  As
> follows:
> >From (gnome terminal: 3751):
> Warning **: No handler for control sewujence 'device-control-string'
> defined(this was repeater a number of times)>20
> Warning **: Attempt to set invalid NRC map I(repeated 2 times)
> Warning **:[invalid UTYF-8] invalid NRC map I (1 time)
> Warning **: Attempt to set invalid  NRC map I (1 time)
> I dont know what happened to k3b it worked previously.
> 
> Also the splash screen (2.4) will not go away by itself but must be
> clicked to remove.

That's not good. Can you show me the output from uptime, or the first
line from top, e.g.

 11:10am  up 40 days, 22 min,  1 user,  load average: 0.08, 0.15, 0.16

I've seen cases where zombie processes have pushed the load average to 4
or 5, and though they don't seem to be using CPU they slow the machine
down (in that case it was the automounter which I don't think is the
problem).

Another thing to try is to create a fresh user (using useradd as root)
and login as that user and see if you get the same behaviour.


Regards,

Chris
-----------------------------------------------------------------
Chris Rouch
crouch pobox com