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/
[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/
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
[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/
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
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
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
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/
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
From: "David L" idht4n@hotmail.com Reply-To: For users of Fedora Core releases fedora-list@redhat.com To: fedora-list@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
_________________________________________________________________ Dont just search. Find. Check out the new MSN Search! http://search.msn.click-url.com/go/onm00200636ave/direct/01/
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
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
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
On 6/3/06, Christopher K. Johnson ckjohnson@gwi.net 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@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
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/
>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
_________________________________________________________________ Dont just search. Find. Check out the new MSN Search! http://search.msn.click-url.com/go/onm00200636ave/direct/01/
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/
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.
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:tonynelson@georgeanelson.com ' http://www.georgeanelson.com/
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/
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
_________________________________________________________________ Dont just search. Find. Check out the new MSN Search! http://search.msn.click-url.com/go/onm00200636ave/direct/01/
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
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
[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/
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
_________________________________________________________________ Dont just search. Find. Check out the new MSN Search! http://search.msn.click-url.com/go/onm00200636ave/direct/01/
-
Christopher K. Johnson -
David -
Jason Anderson -
Jay Cliburn -
Jim Cornette -
Keith G. Robertson-Turner -
Paul Dickson -
Randy Wyatt -
Ric Moore -
Tony Nelson