fc5 hangs

[David]

Two days in a row my fc5 computer has frozen soon after I start using it in 
the morning.  The screen wakes up from screensaver mode when I move the 
mouse and I'm able to do a few things like change the viewport... then the 
screen goes blank and all I can see is the outline of the mouse pointer 
which is no longer responsive to mouse motion.  The keyboard is unresponsive 
to control-alt-backspace and control-alt-delete and caps lock and num lock 
fail to change the keyboard lights.  How do I attempt to identify the 
problem in a case like this?  I think I might have clicked on an evolution 
send/receive button soon before both crashes, so my only guess right now is 
an evolution problem.

             David

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

[David]

[snip]
> Two days in a row my fc5 computer has frozen soon after I start using it in 
> the morning.  The screen wakes up from screensaver mode when I move the 
> mouse and I'm able to do a few things like change the viewport... then the 
> screen goes blank and all I can see is the outline of the mouse pointer 
> which is no longer responsive to mouse motion.  The keyboard is 
> unresponsive to control-alt-backspace and control-alt-delete and caps lock 
> and num lock fail to change the keyboard lights.  How do I attempt to 
> identify the problem in a case like this?  I think I might have clicked on 
> an evolution send/receive button soon before both crashes, so my only guess 
> right now is an evolution problem. 
It's 4 of 5 work days now that fc5 has hung on me first thing in the 
morning.
Before today's hang, I had changed my desktop environment from KDE to
GNOME and had changed the screensaver to blank-screen-only, so KDE
and screensaves are not the problem.

Evolution still seems like the common thread, but I have no alternative way
to get email from my company's exchange server, so I can't really experiment
with not using evolution.  I had uptimes exceeding 180 days with fc3 before
upgrading to fc5.  Now my maximum uptime is 3 days (2 of which were
weekend days when I didn't use it).  If anybody has a practical
workaround for this problem, I'd appreciate it.  Or if anybody can suggest
a way to debug keyboard/mouse hangs, I can try to provide a more useful
description of the problem to the list.

Thanks...

          David

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

[Jim Cornette]

David L wrote:
>  [snip]
> >
> > Two days in a row my fc5 computer has frozen soon after I start using 
> > it in the morning.  The screen wakes up from screensaver mode when I 
> > move the mouse and I'm able to do a few things like change the 
> > viewport... then the screen goes blank and all I can see is the 
> > outline of the mouse pointer which is no longer responsive to mouse 
> > motion.  The keyboard is unresponsive to control-alt-backspace and 
> > control-alt-delete and caps lock and num lock fail to change the 
> > keyboard lights.  How do I attempt to identify the problem in a case 
> > like this?  I think I might have clicked on an evolution send/receive 
> > button soon before both crashes, so my only guess right now is an 
> > evolution problem.
>  
>  It's 4 of 5 work days now that fc5 has hung on me first thing in the 
>  morning.
>  Before today's hang, I had changed my desktop environment from KDE to
>  GNOME and had changed the screensaver to blank-screen-only, so KDE
>  and screensaves are not the problem.
>  
>  Evolution still seems like the common thread, but I have no alternative way
>  to get email from my company's exchange server, so I can't really 
>  experiment
>  with not using evolution.  I had uptimes exceeding 180 days with fc3 before
>  upgrading to fc5.  Now my maximum uptime is 3 days (2 of which were
>  weekend days when I didn't use it).  If anybody has a practical
>  workaround for this problem, I'd appreciate it.  Or if anybody can suggest
>  a way to debug keyboard/mouse hangs, I can try to provide a more useful
>  description of the problem to the list.
>  
>  Thanks...
>  
>           David
>   
You might try installing the synaptics mouse driver from the repos. 
After it is installed, it might help if you run
system-config-display --reconfig
after backing up your current xorg.conf file and after the package is 
installed.

If you have any information in your kernel command related to the mouse 
working, remove it and reboot the computer before running.
system-config-display --reconfig

It might be helpful if you could give the output for your hardware for 
videocard, keyboard, mouse and the like.

Jim

-- 
Many people are unenthusiastic about your work.

[David]

[snip]
> >>Two days in a row my fc5 computer has frozen soon after I
> start using it 
> >>in the morning. 
[snip]
> You might try installing the synaptics mouse driver from the repos. 
It appeared to be installed already.  yum install synaptics said there
was nothing to do.

> After it is installed, it might help if you run
> system-config-display --reconfig 
I tried that, but not until after today's
hang.

However, I have more information about the problem.  I think I misdiagnosed 
the problem...
it appears that the mouse/keyboard are not completely unresponsive... 
they're are just
responding extremely slowly.  I ssh'd into my work computer (the one that is 
"hanging"
every day) from home last night.  This morning before I came to work, I 
tried to do a little
work from my ssh session.  And it was messed up too!  Things were running 
very slowly.
I tried to run top and it took a few minutes to even start.  Once it 
started, it was usually
quite responsive (even when a second ssh session was still sluggish) but 
sometimes
stopped responding for ~10 seconds.  When it was running, it didn't show any 
process
hogging the CPU and the load average was only around 2-3.  There was not a 
whole lot
of free memory, but I freed up some memory by killing a few processes and 
the problem
didn't go away.  I killed evolution, evolution-exchange, etc.  I tried to 
run a command in
my (autofs mounted) home directory and it gave some error that I 
unfortunately can't
remember.  I restarted autofs (which took about 5 minutes) and I was able to 
access my
home directory again, but the system was still sluggish.  I noticed that the 
clock was ~3
hours behind despite the fact that ntpd was running.  I also noticed that 
some daily cron
stuff was running (namely prelink).  I stopped crond and killed everything 
that crond had
started.  It was still extremely slow.  Finally, I came to work and tried to 
interact with it
locally and saw similar behavior to what I see about once per day.  I tried 
to log out
from  my X session to see if that would help, but I only waited 10 minutes 
for it to log out.
It did seem to be slowly closing windows and logging out, so it might have 
worked if I
waited an hour, but I finally got impatient and pushed the reset button.

Any thoughts on what could be causing this behavior?  It acts like the CPU 
is overloaded
or we're out of memory and swap, but that is not the case.

Thanks...

               David

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

[Jim Cornette]

David L wrote:
>  [snip]
> >>> Two days in a row my fc5 computer has frozen soon after I start 
> >>> using it in the morning.
>  [snip]
> >
> > You might try installing the synaptics mouse driver from the repos.
>  
>  It appeared to be installed already.  yum install synaptics said there
>  was nothing to do.
>  
> > After it is installed, it might help if you run
> > system-config-display --reconfig
>  I tried that, but not until after today's hang. 
I was hoping something simple like re-configuring X would solve the problem.

>  
>  However, I have more information about the problem.  I think I 
>  misdiagnosed the problem...
>  it appears that the mouse/keyboard are not completely unresponsive... 
>  they're are just
>  responding extremely slowly.  I ssh'd into my work computer (the one 
>  that is "hanging"
>  every day) from home last night.  This morning before I came to work, I 
>  tried to do a little
>  work from my ssh session.  And it was messed up too!  Things were 
>  running very slowly. 
I would guess networking problems, maybe ipv6 or whatever the protocol 
is called. Since no process seems to be hogging memory or swap.

>  I tried to run top and it took a few minutes to even start.  Once it
>
>  started, it was usually
>  quite responsive (even when a second ssh session was still sluggish) but 
>  sometimes
>  stopped responding for ~10 seconds.  When it was running, it didn't show 
>  any process
>  hogging the CPU and the load average was only around 2-3.  There was not 
>  a whole lot
>  of free memory, but I freed up some memory by killing a few processes 
>  and the problem
>  didn't go away.  I killed evolution, evolution-exchange, etc.  
The beagle daemon was said to hog process power.

  I tried
>  to run a command in
>  my (autofs mounted) home directory and it gave some error that I 
>  unfortunately can't
>  remember.  I restarted autofs (which took about 5 minutes) and I was 
>  able to access my
>  home directory again, but the system was still sluggish.  I noticed that 
>  the clock was ~3
>  hours behind despite the fact that ntpd was running.  I also noticed 
>  that some daily cron
>  stuff was running (namely prelink).  I stopped crond and killed 
>  everything that crond had
>  started.  It was still extremely slow.  
Prelink slowed down my system on occasion previously. The program seems 
to be pretty much in control now and does not bog down my system.

  Finally, I came to work and
>  tried to interact with it
>  locally and saw similar behavior to what I see about once per day.  I 
>  tried to log out
>  from  my X session to see if that would help, but I only waited 10 
>  minutes for it to log out.
>  It did seem to be slowly closing windows and logging out, so it might 
>  have worked if I
>  waited an hour, 
This might indicate a network bottleneck. Enough wild guesses. :-)

  but I finally got impatient and pushed the reset button.
>  
>  Any thoughts on what could be causing this behavior?  It acts like the 
>  CPU is overloaded
>  or we're out of memory and swap, but that is not the case. 
No real clue what is going on. My computer does this with the program 
wine. Other programs do not kill X, nautilus and such.

Exit, stage left,
Jim

>  
>  Thanks...
>  
>                David 
-- 
Our policy is, when in doubt, do the right thing.
		-- Roy L. Ash, ex-president, Litton Industries

[Jason Anderson]

David,

It sounds like something hardware-related.  Try updating your kernel
first, if you're not using the latest one.  I recently had a similar
problem that was caused by a kernel bug relating to Athlon64 X2
processors.  It's fixed now.

Have you recently added or changed your hardware?  Run 'dmesg|less' and
look near the bottom for errors.

If that's not it, take out or unplug each piece of hardware and retest
until all your peripherals are out.  If that doesn't help, find out
which video driver you are using, and try to find a temporary
alternative.  That might tell you if your video driver or card is the
problem.

jason

On Tue, 2006-05-23 at 21:53 -0400, Jim Cornette
fc-cornette-at-insight.rr.com |fedora-list| wrote:
>  David L wrote:
>  > [snip]
>  >>>> Two days in a row my fc5 computer has frozen soon after I start 
>  >>>> using it in the morning.
>  > [snip]
>  >>
>  >> You might try installing the synaptics mouse driver from the repos.
>  > 
>  > It appeared to be installed already.  yum install synaptics said there
>  > was nothing to do.
>  > 
>  >> After it is installed, it might help if you run
>  >> system-config-display --reconfig
>  > I tried that, but not until after today's hang.
>  
>  I was hoping something simple like re-configuring X would solve the problem.
>  
>  > 
>  > However, I have more information about the problem.  I think I 
>  > misdiagnosed the problem...
>  > it appears that the mouse/keyboard are not completely unresponsive... 
>  > they're are just
>  > responding extremely slowly.  I ssh'd into my work computer (the one 
>  > that is "hanging"
>  > every day) from home last night.  This morning before I came to work, I 
>  > tried to do a little
>  > work from my ssh session.  And it was messed up too!  Things were 
>  > running very slowly.
>  
>  I would guess networking problems, maybe ipv6 or whatever the protocol 
>  is called. Since no process seems to be hogging memory or swap.
>  
>  > I tried to run top and it took a few minutes to even start.  Once it 
>  > started, it was usually
>  > quite responsive (even when a second ssh session was still sluggish) but 
>  > sometimes
>  > stopped responding for ~10 seconds.  When it was running, it didn't show 
>  > any process
>  > hogging the CPU and the load average was only around 2-3.  There was not 
>  > a whole lot
>  > of free memory, but I freed up some memory by killing a few processes 
>  > and the problem
>  > didn't go away.  I killed evolution, evolution-exchange, etc. 
>  
>  The beagle daemon was said to hog process power.
>  
>    I tried
>  > to run a command in
>  > my (autofs mounted) home directory and it gave some error that I 
>  > unfortunately can't
>  > remember.  I restarted autofs (which took about 5 minutes) and I was 
>  > able to access my
>  > home directory again, but the system was still sluggish.  I noticed that 
>  > the clock was ~3
>  > hours behind despite the fact that ntpd was running.  I also noticed 
>  > that some daily cron
>  > stuff was running (namely prelink).  I stopped crond and killed 
>  > everything that crond had
>  > started.  It was still extremely slow. 
>  
>  Prelink slowed down my system on occasion previously. The program seems 
>  to be pretty much in control now and does not bog down my system.
>  
>    Finally, I came to work and
>  > tried to interact with it
>  > locally and saw similar behavior to what I see about once per day.  I 
>  > tried to log out
>  > from  my X session to see if that would help, but I only waited 10 
>  > minutes for it to log out.
>  > It did seem to be slowly closing windows and logging out, so it might 
>  > have worked if I
>  > waited an hour,
>  
>  This might indicate a network bottleneck. Enough wild guesses. :-)
>  
>    but I finally got impatient and pushed the reset button.
>  > 
>  > Any thoughts on what could be causing this behavior?  It acts like the 
>  > CPU is overloaded
>  > or we're out of memory and swap, but that is not the case.
>  
>  No real clue what is going on. My computer does this with the program 
>  wine. Other programs do not kill X, nautilus and such.
>  
>  Exit, stage left,
>  Jim
>  
>  > 
>  > Thanks...
>  > 
>  >               David
>  
>  -- 
>  Our policy is, when in doubt, do the right thing.
>  		-- Roy L. Ash, ex-president, Litton Industries
>

[David]

> >>>>Two days in a row my fc5 computer has frozen soon
> after I start using 
> >>>>it in the morning. 
[snip]

> >However, I have more information about the problem.  I think I 
> >misdiagnosed the problem...
> >it appears that the mouse/keyboard are not completely unresponsive... 
> >they're are just
> >responding extremely slowly.  I ssh'd into my work computer (the one that 
> >is "hanging"
> >every day) from home last night.  This morning before I came to work, I 
> >tried to do a little
> >work from my ssh session.  And it was messed up too!  Things were running 
> >very slowly. 
[snip]

My fc5 system hasn't "hung" in 5 days, so I'm hoping that something
that I've tried has fixed it.  Still too soon to tell for sure I guess, but 
here's
what I've tried:

yum -y update
(It was already up to date as of about 10 days ago, but I updated again).

Disabled the following services:
acpid
apmd
avahi-daemon
bluetooth
cpuspeed
hidd
irqbalance
isdn

Removed the following from cron.daily:
prelink
beagle-crawl-system

I'm hoping that one of these things fixed the problem.

          David

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

[Ric Moore]

On Sun, 2006-05-28 at 08:16 -0700, David L wrote:

>  >>>>>Two days in a row my fc5 computer has frozen soon
> after I start using 
>  >>>>>it in the morning.
>  
>  [snip]
>  
>  >>However, I have more information about the problem.  I think I 
>  >>misdiagnosed the problem...
>  >>it appears that the mouse/keyboard are not completely unresponsive... 
>  >>they're are just
>  >>responding extremely slowly.  I ssh'd into my work computer (the one that 
>  >>is "hanging"
>  >>every day) from home last night.  This morning before I came to work, I 
>  >>tried to do a little
>  >>work from my ssh session.  And it was messed up too!  Things were running 
>  >>very slowly.
>  
>  [snip]
>  
>  My fc5 system hasn't "hung" in 5 days, so I'm hoping that something
>  that I've tried has fixed it.  Still too soon to tell for sure I guess, but 
>  here's
>  what I've tried:
>  
>  yum -y update
>  (It was already up to date as of about 10 days ago, but I updated again).
>  
>  Disabled the following services:
>  acpid
>  apmd
>  avahi-daemon
>  bluetooth
>  cpuspeed
>  hidd
>  irqbalance
>  isdn
>  
>  Removed the following from cron.daily:
>  prelink
>  beagle-crawl-system
>  
>  I'm hoping that one of these things fixed the problem. 
Yeah,  Beagle started eating up CPU this morning. It practically stopped
my system cold. <Sigh> Ric

[David]

> From: "David L" <idht4n(a)hotmail.com&gt;
> Reply-To: For users of Fedora Core releases <fedora-list(a)redhat.com&gt;
> To: fedora-list(a)redhat.com
> Subject: Re: fc5 hangs
> Date: Sun, 28 May 2006 08:16:10 -0700
>
> >>>>>Two days in a row my fc5 computer has frozen soon after I start using
>
> >>>>>it in the morning.
>
> [snip]
>
> >>However, I have more information about the problem.  I think I 
> >>misdiagnosed the problem...
> >>it appears that the mouse/keyboard are not completely unresponsive... 
> >>they're are just
> >>responding extremely slowly.  I ssh'd into my work computer (the one that 
> >>is "hanging"
> >>every day) from home last night.  This morning before I came to work, I 
> >>tried to do a little
> >>work from my ssh session.  And it was messed up too!  Things were running 
> >>very slowly.
>
> [snip]
>
> My fc5 system hasn't "hung" in 5 days, so I'm hoping that something
> that I've tried has fixed it. 
Arg!  It's still messed up.  Same weird symptoms.  Very slow to log in
and run commands, but usually quite reponsive when things are running.
For example, 90% of the time, a shell is responsive when I'm typing (the
other 10% of the time it's completely unresponsive for ~5 seconds).  But
when I run a command like top, it takes ~30 seconds to run.  Then when
top is running, it's pretty responsive 90% of the time and shows low load
averages.  The only weird thing I notice in top is that I sometimes see
-0 CPU usage on some processes.  Here's an example top output
when the system is screwed up:

top - 17:53:52 up  8:32,  9 users,  load average: 1.39, 0.53, 0.19
Tasks: 149 total,   2 running, 144 sleeping,   2 stopped,   1 zombie
Cpu(s):  0.0% us,  0.0% sy,  0.0% ni,  0.0% id,  0.0% wa,  0.0% hi,  0.0% 
si,  0.0% st
Mem:   2067636k total,  1185580k used,   882056k free,    47392k buffers
Swap:  6144820k total,        0k used,  6144820k free,   661912k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
    1 root      16   0  1992  684  584 S -0.0  0.0   0:00.69 init
    2 root      34  19     0    0    0 S -0.0  0.0   0:00.20 ksoftirqd/0
    3 root      RT   0     0    0    0 S -0.0  0.0   0:00.00 watchdog/0
    4 root      10  -5     0    0    0 S -0.0  0.0   0:00.28 events/0
    5 root      10  -5     0    0    0 S -0.0  0.0   0:00.00 khelper
    6 root      10  -5     0    0    0 S -0.0  0.0   0:00.00 kthread
    8 root      10  -5     0    0    0 S -0.0  0.0   0:00.01 kblockd/0
    9 root      20  -5     0    0    0 S -0.0  0.0   0:00.00 kacpid
   96 root      10  -5     0    0    0 S -0.0  0.0   0:00.00 khubd
  151 root      20   0     0    0    0 S -0.0  0.0   0:00.00 pdflush
  152 root      15   0     0    0    0 S -0.0  0.0   0:00.01 pdflush
  154 root      11  -5     0    0    0 S -0.0  0.0   0:00.00 aio/0
  153 root      25   0     0    0    0 S -0.0  0.0   0:00.00 kswapd0
  241 root      10  -5     0    0    0 S -0.0  0.0   0:00.00 kseriod
  319 root      11  -5     0    0    0 S -0.0  0.0   0:00.00 kpsmoused
  332 root      15   0     0    0    0 S -0.0  0.0   0:00.06 kjournald
  371 root      10  -5     0    0    0 S -0.0  0.0   0:00.00 kauditd
  395 root      12  -4  2208  692  376 S -0.0  0.0   0:00.09 udevd
1170 root      11  -5     0    0    0 S -0.0  0.0   0:00.00 scsi_eh_0
1171 root      10  -5     0    0    0 S -0.0  0.0   0:00.79 usb-storage
1288 root      11  -5     0    0    0 S -0.0  0.0   0:00.00 kmpathd/0
1295 root      11  -5     0    0    0 S -0.0  0.0   0:00.00 kmirrord
1317 root      23   0     0    0    0 S -0.0  0.0   0:00.00 kjournald
1600 root      13  -3 11988  620  468 S -0.0  0.0   0:00.02 auditd
1613 root      16   0  1652  552  460 S -0.0  0.0   0:00.04 syslogd
1616 root      16   0  1604  396  328 S -0.0  0.0   0:00.00 klogd
1628 rpc       16   0  1732  600  504 S -0.0  0.0   0:00.01 portmap
1647 rpcuser   16   0  1784  760  656 S -0.0  0.0   0:00.00 rpc.statd
1670 root      16   0  4728  680  336 S -0.0  0.0   0:00.02 rpc.idmapd
1684 dbus      16   0  3108 1284 1028 S -0.0  0.1   0:01.00 dbus-daemon
1717 root      20   0 24312  436  300 S -0.0  0.0   0:00.00 ypbind
1779 root      16   0  1872  752  612 S -0.0  0.0   0:00.00 automount
1808 root      16   0  1884  828  676 S -0.0  0.0   0:00.00 automount
1832 root      16   0  1884  836  676 S -0.0  0.0   0:00.00 automount
1861 root      16   0  1880  816  676 S -0.0  0.0   0:00.00 automount
1876 root      16   0  1896  488  288 S -0.0  0.0   0:00.00 smartd
1885 root      15   0  5004  484  328 S -0.0  0.0   0:00.00 hpiod
1890 root      16   0 13056 5220 1172 S -0.0  0.3   0:00.06 python
1901 root      16   0  9180 2660 1944 S -0.0  0.1   0:00.05 cupsd
1909 root      16   0  4980 1112  788 S -0.0  0.1   0:00.00 sshd
1919 root      15   0  2232  816  676 S -0.0  0.0   0:00.00 xinetd
1931 ntp       16   0  4272 4272 3272 S -0.0  0.2   0:00.05 ntpd
1949 root      16   0  8284 1908  932 S -0.0  0.1   0:00.01 sendmail
1956 smmsp     16   0  7344 1728  884 S -0.0  0.1   0:00.00 sendmail
1966 root      16   0  1820  344  280 S -0.0  0.0   0:00.04 gpm
1976 root      16   0  5176 1112  572 S -0.0  0.1   0:00.17 crond
2011 xfs       16   0  6224 4204  904 S -0.0  0.2   0:00.51 xfs
2028 root      16   0  2164  452  324 S -0.0  0.0   0:00.00 atd
2091 root      24   0  3132 1160 1048 S -0.0  0.1   0:00.00 cups-config-dae
2101 haldaemo  16   0  5024 3368 1760 S -0.0  0.2   0:01.74 hald
2102 root      16   0  3136 1024  900 S -0.0  0.0   0:00.01 hald-runner
2108 haldaemo  16   0  2232  856  768 S -0.0  0.0   0:00.00 hald-addon-acpi
2117 haldaemo  15   0  2232  868  776 S -0.0  0.0   0:00.59 hald-addon-keyb
2121 root      16   0  2196  704  624 S -0.0  0.0   0:00.52 hald-addon-stor
2124 root      16   0  2196  704  624 S -0.0  0.0   0:00.32 hald-addon-stor
2126 root      16   0  2196  704  624 S -0.0  0.0   0:00.53 hald-addon-stor
2149 root      17   0  1588  408  348 S -0.0  0.0   0:00.00 mingetty
2160 root      17   0  1584  404  348 S -0.0  0.0   0:00.00 mingetty
2163 root      17   0  1588  408  348 S -0.0  0.0   0:00.00 mingetty
2166 root      17   0  1588  404  348 S -0.0  0.0   0:00.00 mingetty
2169 root      17   0  1588  408  348 S -0.0  0.0   0:00.00 mingetty
2172 root      18   0  1588  408  348 S -0.0  0.0   0:00.00 mingetty
2183 root      18   0  4448 1232 1068 S -0.0  0.1   0:00.00 prefdm
2188 root      16   0 12728 2488 2052 S -0.0  0.1   0:00.02 gdm-binary
2245 root      16   0 13388 2996 2272 S -0.0  0.1   0:00.02 gdm-binary
2248 root      16   0  174m  34m  11m S -0.0  1.7   7:53.54 Xorg
2322 root      10  -5     0    0    0 S -0.0  0.0   0:08.00 rpciod/0
2323 root      19   0     0    0    0 S -0.0  0.0   0:00.00 lockd

Any clues in there?  Any suggestions on how to debug a system that
has low load averages, plenty of free memory, but yet is unusably slow?

Thanks...

           David

_________________________________________________________________
Don’t just search. Find. Check out the new MSN Search! 
http://search.msn.click-url.com/go/onm00200636ave/direct/01/

[David]

> Two days in a row my fc5 computer has frozen soon after I start using
> it in 
> the morning.
> [snip]
>
> >However, I have more information about the problem.  I think I 
> >misdiagnosed the problem...
> >it appears that the mouse/keyboard are not completely unresponsive... 
> >they're are just
> >responding extremely slowly.  I ssh'd into my work computer (the one that 
> >is "hanging"
> >every day) from home last night.  This morning before I came to work, I 
> >tried to do a little
> >work from my ssh session.  And it was messed up too!  Things were running 
> >very slowly.
> >
> >[snip]
> >
> >My fc5 system hasn't "hung" in 5 days, so I'm hoping that something
> >that I've tried has fixed it.
>
> Arg!  It's still messed up.  Same weird symptoms.  Very slow to log in 
I recovered from this state without rebooting this time.  I
started killing processes and stopping services and finally
I did a telinit 3.   After all that, the system was responsive
again.  I killed all of the processes owned by me (so I could
stop autofs which automounts my home directory), then I
stopped autofs, ypbind, smartd, haldaemon, ???, and then
/sbin/telinit 3.  I'm not sure exactly when the system became
happy again, but I think it was after telinit.  I'll try doing only
that next time it happens and post the results.

           David

_________________________________________________________________
Is your PC infected? Get a FREE online computer virus scan from McAfee® 
Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

[David]

> >Two days in a row my fc5 computer has frozen soon after I start
> using it 
> >in the morning. 
Make it about 80% of the days since I installed fc5.

> >
> >>However, I have more information about the problem.  I think I 
> >>misdiagnosed the problem...
> >>it appears that the mouse/keyboard are not completely unresponsive... 
> >>they're are just
> >>responding extremely slowly.  I ssh'd into my work computer (the one that 
> >>is "hanging"
> >>every day) from home last night.  This morning before I came to work, I 
> >>tried to do a little
> >>work from my ssh session.  And it was messed up too!  Things were running 
> >>very slowly.
> >>
> >>[snip]
> >>
> >>My fc5 system hasn't "hung" in 5 days, so I'm hoping that
> something
> >>that I've tried has fixed it.
> >
> >Arg!  It's still messed up.  Same weird symptoms.  Very slow to log in
>
> I recovered from this state without rebooting this time.  I
> started killing processes and stopping services and finally
> I did a telinit 3.   After all that, the system was responsive
> again.  I killed all of the processes owned by me (so I could
> stop autofs which automounts my home directory), then I
> stopped autofs, ypbind, smartd, haldaemon, ???, and then
> /sbin/telinit 3.  I'm not sure exactly when the system became
> happy again, but I think it was after telinit.  I'll try doing only
> that next time it happens and post the results. 
telinit 3 by itself didn't help.  Can anybody suggest how to
debug this?  I've been using redhat/fedora for ~11 years now,
but I'm about to jump ship.  I'm losing way too much productivity
with this bug.  :(

Thanks...

                  David

_________________________________________________________________
On the road to retirement? Check out MSN Life Events for advice on how to 
get there! http://lifeevents.msn.com/category.aspx?cid=Retirement

[Christopher K. Johnson]

David L wrote:
>  telinit 3 by itself didn't help.  Can anybody suggest how to
>  debug this?  I've been using redhat/fedora for ~11 years now,
>  but I'm about to jump ship.  I'm losing way too much productivity
>  with this bug.  :( 
Have you tried the 'top' command to see what processes
are using 
resources?  Without options it auto refreshes so you type 'q' to quit.

Chris

-- 
   "Spend less!  Do more!  Go Open Source..." -- Dirigo.net
   Chris Johnson, RHCE #804005699817957

[Randy Wyatt]

On 6/3/06, Christopher K. Johnson <ckjohnson(a)gwi.net&gt; wrote:
>  David L wrote:
>  >
>  >
>  > telinit 3 by itself didn't help.  Can anybody suggest how to
>  > debug this?  I've been using redhat/fedora for ~11 years now,
>  > but I'm about to jump ship.  I'm losing way too much productivity
>  > with this bug.  :(
>  Have you tried the 'top' command to see what processes are using
>  resources?  Without options it auto refreshes so you type 'q' to quit.
>
>  Chris
>
>  --
>     "Spend less!  Do more!  Go Open Source..." -- Dirigo.net
>     Chris Johnson, RHCE #804005699817957
>
>  --
>  fedora-list mailing list
>  fedora-list(a)redhat.com
>  To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
>  
Out of curiousity, I wonder if a web browser is being left open all
night.  Firefox tends to eat a good amount of memory

[David]

> >
> >telinit 3 by itself didn't help.  Can anybody suggest how to
> >debug this?  I've been using redhat/fedora for ~11 years now,
> >but I'm about to jump ship.  I'm losing way too much productivity
> >with this bug.  :(
> Have you tried the 'top' command to see what processes are using resources? 
>   Without options it auto refreshes so you type 'q' to quit.
>  
Yes, I've tried top... I posted the results earlier, so I won't post
them again here.  The load average was low, there was plenty
of memory, and no process seemed to be sucking CPU.  As I
mentioned in that earlier post, it sometimes showed all of the
processes as using -0.0 % of the CPU... I don't know what the
"-" sign was about, but maybe it's a clue.

Thanks...
                David

_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar – get it now! 
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/

[David]

> >>>>>>Two days in a row my fc5 computer has frozen
> soon after I start using 
> >>>>>>it in the morning.
> >
> >[snip]
> >
> >>>However, I have more information about the problem.  I think I 
> >>>misdiagnosed the problem...
> >>>it appears that the mouse/keyboard are not completely unresponsive... 
> >>>they're are just
> >>>responding extremely slowly. 
I think I've discovered an important clue.?  I noticed the time
was frequently way off when this problem occurred, so I tried
cat /proc/interrupts | grep timer

The timer interrupt is incrementing very slowly.  Instead of
1000 Hz it's incrementing about once or twice per minute.  What
would cause that?  I had ntpd running, but I killed it.

Thanks...
                    David

_________________________________________________________________
Don’t just search. Find. Check out the new MSN Search! 
http://search.msn.click-url.com/go/onm00200636ave/direct/01/

timer broken (was Re: fc5 hangs)

[David]

> >>Two days in a row my fc5 computer has frozen soon after I
> start using it 
> >>in the morning.
> >However, I have more information about the problem.  I think I 
> >misdiagnosed the problem...
> >it appears that the mouse/keyboard are not completely unresponsive... 
> >they're are just
> >responding extremely slowly.
>
> I think I've discovered an important clue.?  I noticed the time
> was frequently way off when this problem occurred, so I tried
> cat /proc/interrupts | grep timer
>
> The timer interrupt is incrementing very slowly.  Instead of
> 1000 Hz it's incrementing about once or twice per minute.  What
> would cause that?  I had ntpd running, but I killed it.
>  
The overnight average for my fc5 box was one timer
interrupt every ~32 seconds.  I left it running in case
somebody posted a diagnostic for me to try while it
was in this state, but no responses, so I rebooted it
this morning so I can start getting some work done.
It took about 10 minutes to reboot, but after it rebooted,
it was operating normally again (1000 Hz timer
interrupts) and good responsiveness.  Any thoughts
on why my timer seems to change to about 1/2^15
slower interrupt rate than usual?  IIRC, this system
is pretty much vanilla fc5.  I haven't installed any
RPMs that aren't from fedora updates or extras.

Thanks...

           David

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

timer broken (was Re: fc5 hangs)

[Keith G. Robertson-Turner]

David L wrote:

> > The timer interrupt is incrementing very slowly.  Instead of
> 1000
> > Hz it's incrementing about once or twice per minute.  What would
> > cause that?  I had ntpd running, but I killed it. 
There's your answer - kernel bug specific to your hardware.

Try downgrading your kernel to an earlier release, and see if the timer
(and other issues) disappear.

Incidentally, during a recent kernel compile, I noticed that 1000Hz is 
not the default setting in the config; I think it's 250Hz IIRC, for some 
reason.

I usually avoid kernel updates completely if all the hardware I have
already works with the current kernel. It also means I can avoid
rebooting and losing uptime. At the very least, I keep the previous
kernel for weeks, maybe months, to be sure I don't fall prey to an
upgrade gotcha.

Beats me why features that work in previous releases of the kernel,
don't work in subsequent versions. You'd have thought it should be the
other way round. Stuff like that, seems to happen only too often with 
the kernel.

Anyway, have a go - and good luck.

-- 
K.
http://slated.org - Slated, Rated & Blogged

Fedora Core release 5 (Bordeaux) on sky, running kernel 2.6.16-1.2122_FC5
  02:00:16 up 8 days, 21:40,  1 user,  load average: 0.13, 0.04, 0.01

timer broken (was Re: fc5 hangs)

[Tony Nelson]

At 2:03 AM +0100 6/5/06, Keith G. Robertson-Turner wrote:
 ...
> Incidentally, during a recent kernel compile, I noticed that 1000Hz
> is
> not the default setting in the config; I think it's 250Hz IIRC, for some
> reason. 
 ...

Laptop power consumption.  At 1000 Hz, laptops often don't get to
power-save between timer interrupts.  Saw it on something (?) related to
lkml.
____________________________________________________________________
TonyN.:'                       <mailto:[e-mail]>
      '                              <http://www.georgeanelson.com/>

timer broken (was Re: fc5 hangs)

[David]

> David L wrote:
>
> >>The timer interrupt is incrementing very slowly.  Instead of 1000
> >>Hz it's incrementing about once or twice per minute.  What would
> >>cause that?  I had ntpd running, but I killed it.
>
> There's your answer - kernel bug specific to your hardware.
>
> Try downgrading your kernel to an earlier release, and see if the timer
> (and other issues) disappear. 
OK, I downgraded to 2.6.12 and let it run overnight.  Oddly, the timer
was getting 4 interrupts per second this morning.  The overnight
average was about 20 interrupts per second, which might mean
it ran for 7.5 minutes at 1000 Hz and then switched to 4 Hz.  However,
the date/time reported is correct, unlike when my system was "hanging"
before.

The system seems pretty fast and responsive, but I think top is
under-reporting the CPU usage by different processes.  Processes
that usually suck a lot of CPU are reporting less than 1%.  Are there
any kernel compatibility compatibility problems with running 2.6.12
with fc5?

Thanks...

                         David

PS - Since my problem was intermittent, I'm not sure if it has gone away.
I still am seeing something weird wrt the timer interrupts.

_________________________________________________________________
FREE pop-up blocking with the new MSN Toolbar – get it now! 
http://toolbar.msn.click-url.com/go/onm00200415ave/direct/01/

timer broken (was Re: fc5 hangs)

[David]

> >David L wrote:
> >
> >>>The timer interrupt is incrementing very slowly.  Instead of 1000
> >>>Hz it's incrementing about once or twice per minute.  What would
> >>>cause that?  I had ntpd running, but I killed it.
> >
> >There's your answer - kernel bug specific to your hardware.
> >
> >Try downgrading your kernel to an earlier release, and see if the timer
> >(and other issues) disappear.
>
> OK, I downgraded to 2.6.12 and let it run overnight. 
[snip]

When I got to work, I found that X-windows had failed to start after
downgrading to the 2.6.12 kernel.  So I tried 2.6.14 and noticed that
udev failed to start.  So I had to bring the kernel back to 2.6.16.  I
compiled 2.6.16.20 without ACPI and APM support and I'm trying
that now.  I guess I'll have to learn how to downgrade udev before
trying an older kernel.

Is the timer behavior I'm seeing definitely a kernel bug?  Or should
I continue my effort of trying older kernels before filing a bugzilla
report on the fc5 kernel?

Thanks...

               David

_________________________________________________________________
Don’t just search. Find. Check out the new MSN Search! 
http://search.msn.click-url.com/go/onm00200636ave/direct/01/

timer broken (was Re: fc5 hangs)

[David]

[snip]
> >>>>The timer interrupt is incrementing very slowly. 
> Instead of 1000
> >>>>Hz it's incrementing about once or twice per minute.  What would
> >>>>cause that?  I had ntpd running, but I killed it.
> >>
> >>There's your answer - kernel bug specific to your hardware.
> >>
> >>Try downgrading your kernel to an earlier release, and see if the timer
> >>(and other issues) disappear.
> >
> >OK, I downgraded to 2.6.12 and let it run overnight.
> [snip]
>
> When I got to work, I found that X-windows had failed to start after
> downgrading to the 2.6.12 kernel.  So I tried 2.6.14 and noticed that
> udev failed to start.  So I had to bring the kernel back to 2.6.16.  I
> compiled 2.6.16.20 without ACPI and APM support and I'm trying
> that now. 
I have not seen the timer problem since I changed to the
2.6.16.20 kernel.  I've filed a bug report against the fc5
kernels:

https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=194914

Thanks to everyone for their suggestions...

David

PS - I'm not sure whether to add "resolved" to the subject line
because this solution is just a workaround.

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/

timer broken (was Re: fc5 hangs)

[David]

> >>David L wrote:
> >>
> >>>>The timer interrupt is incrementing very slowly.  Instead of 1000
> >>>>Hz it's incrementing about once or twice per minute.  What would
> >>>>cause that?  I had ntpd running, but I killed it.
> >>
> >>There's your answer - kernel bug specific to your hardware.
> >>
> >>Try downgrading your kernel to an earlier release, and see if the timer
> >>(and other issues) disappear.
> >
> >OK, I downgraded to 2.6.12 and let it run overnight.
> [snip]
>
> When I got to work, I found that X-windows had failed to start after
> downgrading to the 2.6.12 kernel.  So I tried 2.6.14 and noticed that
> udev failed to start.  So I had to bring the kernel back to 2.6.16.  I
> compiled 2.6.16.20 without ACPI and APM support and I'm trying
> that now. 
Apparently all bugzilla changes in the last week were lost, so the
bug report I filed on this issue has disappeared.  What I learned
so far was that the IBM ThinkCentre had a buggy BIOS, so they asked
me to update it, which I did.  I'm trying to run the default fc5 kernel
now with the updated kernel.  Apparently newer kernels use more
of the ACPI features than older kernels and they suspected that
a buggy BIOS was to blame.

        David

_________________________________________________________________
Don’t just search. Find. Check out the new MSN Search! 
http://search.msn.click-url.com/go/onm00200636ave/direct/01/

[Paul Dickson]

On Tue, 30 May 2006 20:28:09 -0700, David L wrote:

>  top - 17:53:52 up  8:32,  9 users,  load average: 1.39, 0.53, 0.19
>  Tasks: 149 total,   2 running, 144 sleeping,   2 stopped,   1 zombie
>  Cpu(s):  0.0% us,  0.0% sy,  0.0% ni,  0.0% id,  0.0% wa,  0.0% hi,  0.0% 
>  si,  0.0% st
>  Mem:   2067636k total,  1185580k used,   882056k free,    47392k buffers
>  Swap:  6144820k total,        0k used,  6144820k free,   661912k cached
>  
>    PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
>      1 root      16   0  1992  684  584 S -0.0  0.0   0:00.69 init
>      2 root      34  19     0    0    0 S -0.0  0.0   0:00.20 ksoftirqd/0
>      3 root      RT   0     0    0    0 S -0.0  0.0   0:00.00 watchdog/0
>      4 root      10  -5     0    0    0 S -0.0  0.0   0:00.28 events/0
>      5 root      10  -5     0    0    0 S -0.0  0.0   0:00.00 khelper
>      6 root      10  -5     0    0    0 S -0.0  0.0   0:00.00 kthread
>      8 root      10  -5     0    0    0 S -0.0  0.0   0:00.01 kblockd/0
>      9 root      20  -5     0    0    0 S -0.0  0.0   0:00.00 kacpid 
Can you send top output to the list that's sorted by CPU usage (which is
usually the default), not PID.  Your list is cut off, it doesn't show
user processes.

Load average says something is running, but the list you posted barely
shows half the processes and none of the running or stopped ones.

If you are running the latest kernel RPM, you should bugzilla this
situation.

	-Paul

[Paul Dickson]

On Tue, 6 Jun 2006 13:10:07 -0700, Paul Dickson wrote:

>  Can you send top output to the list that's sorted by CPU usage
> (which is
>  usually the default), not PID.  Your list is cut off, it doesn't show
>  user processes.
>  
>  Load average says something is running, but the list you posted barely
>  shows half the processes and none of the running or stopped ones.
>  
>  If you are running the latest kernel RPM, you should bugzilla this
>  situation. 
You seemed to have found your problem focus later in the renamed threaded.

	-Paul

[Jay Cliburn]

On Mon, 2006-05-22 at 10:04 -0700, David L wrote:
>  [snip]
>  >
>  >Two days in a row my fc5 computer has frozen soon after I start using it in 
>  >the morning.  
>  
>  It's 4 of 5 work days now that fc5 has hung on me first thing in the 
>  morning.
>  Before today's hang, I had changed my desktop environment from KDE to
>  GNOME and had changed the screensaver to blank-screen-only, so KDE
>  and screensaves are not the problem.
>  
>  Evolution still seems like the common thread,  
I've seen the same thing since I upgraded to the 2122 kernel on May 22,
and I too seem to have correlation to evolution (since that was the only
app active at the time).  It happened for the first time the evening
after upgrading the kernel (2.6.16-1.2122_FC5.x86_64) on this Athlon64
3000+/Abit AV8 motherboard.  It happened again the next morning (May
23), then again about an hour ago (May 24).  No trace of trouble is left
in the syslog. maillog, or Xlog.  I don't lose the display; the system
just hangs, unresponsive to mouse movement, button clicks, or keyboard
input.  The only recourse is a hard reset.

This has not happened with any previous version of FC5.

Jay