-
Random freezing - no logs
Hello everyone,
I've been trying to find information on how to diagnose my random freezing, but I haven't been very successful. I have tried diagnosing the issue looking through the logs but everything I tried has not been fruitful.
About a month ago I lost a hard drive. It was my OS hard drive. When I replaced it I bought an SSD disk. The original disk was a SATA WD Raptor. Once I reinstalled the OS my system ran fine for a week or two, then the system started freezing and it has been getting worse.
On one of these occasions, I was in the middle of typing and the letter I had pressed at the moment of freezing kept typing. This led me to believe that it was not freezing, it was just the input that were freezing. I tried SSHing to the system form my Windows box using Putty but the connection timed out.
Once it started happening, I checked the logs and nothing was obvious. However, I notice a few things out of the ordinary. One thing was that my USB mouse kept bring up disconnected message and they reconnecting. I pulled on the cable a little bit and I was able to produce these logs by monitoring the logs with tail -f /var/log/messages. I changed out the mouse for another one that did not cause this problem and I checked the logs again and it froze still.
The next thing I tried, since I noticed that one of my drives kept bring up trying to access beyond end of device:
Originally Posted by /var/log/messages
Aug 1 23:57:45 roxbox kernel: [ 6763.045285] attempt to access beyond end of device
Aug 1 23:57:45 roxbox kernel: [ 6763.045287] sdd1: rw=0, want=2150136600, limit=1953519616
This message repeats about 30 times all at once for every instance of the log. I tried unplugging that, and the rest of the drives that were not the SSD OS disk. Still the freezing keeps happening and it seems to be getting worse but the long I use the system the more it seems like it's just random. Sometimes it will freeze in 10 minutes and sometimes it will only freeze once a day.
I don't know where else to look.
-
I would say make sure Plug and Play OS is set to Off in the bios, and perhaps roll back your graphics drivers if you updated them around the time this started. Otherwise google something like "Configure SSD in Linux" because I remember seeing a few little gotchas in there - no show stoppers though....
Need help in realtime? Visit us at #linuxnewbie on irc.libera.chat
Few of us will do as much for our fellow man as he has done.
--Andrew Morton on RMS
-
Thanks je_fro,
I've taken your suggestions into account and this is what else I have done.
I noticed that I left another disk attached to the SATA controller that the SSD is connected to, the other controller has to be disabled as it causes issues with the SSD, it's an older VIA chipset/controller. The controller the disk is connected to is a Promise RAID SATA controller. I disconnected this disk and had not change in behaviour.
I checked the BIOS for the Plug and Play OS and it was indeed set to "NO".
I reset the xorg.conf file back to the original configuration with the nouveau nVidia driver, instead of the proprietary binary nVidia driver that I installed when I installed the SSD disk.
I looked at some guides to configuring SSDs under Linux and what I found seemed to be in regards to lifespan and performance. IE I set the scheduler to deadline in the /etc/lili.conf and then I ran /sbin/lilo. Next I checked "cat /sys/block/sda/queue/scheduler" with the output as follows: noop [deadline] cfq. I set the mount options noatime, nodiratime, discard and I disabled the dump flag. Other than that I didn't find anything else on SSD configuration.
I let the computer run in the default runlevel 3 - text-only mode - and still the system is frozen in the morning.
I've tried everything I can think of. Do you have any other ideas je_fro?
-
How about showing us the output of
The error message of "attempt to access beyond end of device" would suggest the disk space ran out in sdd1?
-
Thanks for your assistance with this issue.
I've run the requested commands in two instances.
Since I've determined that none of the hard drives being disconnected are causing an issue, I ran the commands you requested the output for with them connected and disconnected. I also noticed the first time I ran 'df' that I had allocated 150% of my ram to tmpfs mounts in association with je_fro's suggestion of looking into configuring SSDs with Linux.
Anyway, here is the output to those two commands.
ALL DRIVES DISCONNECTED EXCEPT SSD
Originally Posted by fdisk -l
Disk /dev/sda: 128.0 GB, 128035676160 bytes
255 heads, 63 sectors/track, 15566 cylinders, total 250069680 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x65ca48a0
Device Boot Start End Blocks Id System
/dev/sda1 * 63 234372284 117186111 83 Linux
/dev/sda2 234372285 250069679 7848697+ 82 Linux swap
Originally Posted by df
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/root 115345508 41621092 67865112 39% /
tmpfs 2023620 0 2023620 0% /dev/shm
tmpfs 2023620 0 2023620 0% /tmp
tmpfs 2023620 0 2023620 0% /var/tmp
ALL DRIVES CONNECTED
Originally Posted by fdisk -l
Disk /dev/sda: 128.0 GB, 128035676160 bytes
255 heads, 63 sectors/track, 15566 cylinders, total 250069680 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x65ca48a0
Device Boot Start End Blocks Id System
/dev/sda1 * 63 234372284 117186111 83 Linux
/dev/sda2 234372285 250069679 7848697+ 82 Linux swap
Disk /dev/sdb: 250.1 GB, 250059350016 bytes
255 heads, 63 sectors/track, 30401 cylinders, total 488397168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x2773d71b
Device Boot Start End Blocks Id System
/dev/sdb1 * 63 484375814 242187876 83 Linux
/dev/sdb2 484375815 488392064 2008125 82 Linux swap
Disk /dev/sdc: 160.0 GB, 160040803840 bytes
255 heads, 63 sectors/track, 19457 cylinders, total 312579695 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0xb1eeb1ee
Device Boot Start End Blocks Id System
/dev/sdc1 63 312576704 156288321 83 Linux
Disk /dev/sdd: 1000.2 GB, 1000204886016 bytes
255 heads, 63 sectors/track, 121601 cylinders, total 1953525168 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x8196239c
Device Boot Start End Blocks Id System
/dev/sdd1 * 2048 1953521663 976759808 7 HPFS/NTFS/exFAT
Originally Posted by df -T
Filesystem Type 1K-blocks Used Available Use% Mounted on
/dev/root ext4 115345508 41621068 67865136 39% /
tmpfs tmpfs 688032 26992 661040 4% /dev/shm
tmpfs tmpfs 688032 20 688012 1% /tmp
tmpfs tmpfs 688032 8628 679404 2% /var/tmp
/dev/sdb1 reiserfs 242180444 231820268 10360176 96% /mnt/hd250
/dev/sdc1 reiserfs 156283544 122480316 33803228 79% /mnt/hd160
/dev/sdd1 ext2 961431880 870167368 42426524 96% /mnt/hd1000
You will notice that I added the -T option in the "drives connected" output, this is due to my noticing that the terabyte disk is reported as HPFS/NTFS/exFAT by fdisk and as ext2 - it's ext2 - by df -T. I chose ext2 because there is drivers available for Windows XP for this filesystem which was a requirement at the time of formatting.
[edit]
Originally Posted by saikee
The error message of "attempt to access beyond end of device" would suggest the disk space ran out in sdd1?
No, there are still a few gigs left on that disk. I think that message has something to do with the above mentioned anomaly regarding NTFS and ext2 fs types being detected for the same partition. This anomaly has been around for quite some time, and never gave me any issues before, even when the partition was reported as 100% full with less than a 100 meg left.
I have a workaround for the beyond end of disk error that I've received as it seems to be cause by ktorrent. I am downloading my various softwares to another disk and then moving them to the storage directory which doesn't seem to have any ill effects.
This mages no sense:
Originally Posted by /var/log/messages
Aug 4 00:48:34 roxbox kernel: [ 1629.074337] sdd1: rw=0, want=6995328400, limit=1953519616
This error messages makes sdd1 look like a program trying to access drive space in the 7TB range and is being reported as a 2TB range when it is only a 1TB range. But I don't care about this issue, I plan on getting a 2TB disk, moving everything to that and formatting this 1TB disk to resolve the NTFS/ext2 issue.
Last edited by WrinkledCheese; 08-03-2012 at 11:54 PM.
Reason: answer saikees question about end of disk
-
I myself would not be happy with the sdd1 partitioned with Type 7 for the NTFS filing system and then formatted it to Ext2.
Linux kernel when reading a file does not always used the partition type but it might be needed by other programs for identification purpose. In any case the partition type must match the filing system if you want a trouble free operation.
I would recommend you back up the content of sdd1 first and use cfdisk or fdisk to alter the partition type back to 83 if it has always been a Linux system. However I notice your sdd1 starts the partition at 2048th cylinder whereas all hard disk partitioned by Linux starts at the 63 cylinder as in a MSDOS partition table, which is used by 99.99% hard disks, the first track of 63 sectors is always reserved for the MBR even every operating system and Bios reads only the first sector. This rest of the area is used by boot loaders and special software like the installers and other software that is able to reboot itself.
As far as I know the recent MS Windows like Vista and Windows 7 installers like to start at 2048th cylinder to anticipate a future change. Thus I suspect your sdd1 was used or formatted by a MS WIndows before.
It is possible that you might have to re-partition the hard disk to start from 63th cylinder (done automatically in every Linux partitioning tool).
Does your problem persist if you disconnect the sdd1 ?
Last edited by saikee; 08-04-2012 at 07:40 AM.
-
Yes it does.
I suspect that it became a "NTFS" partition when someone couldn't get it to work on their XP machine with the ext2 filesystem.
I have disconnected sdb through sdd at the same time and the problem still persists. My solution for sdd is to get a hard drive big enough to move everything onto it and then format it/repartition it.
sdd is a 1TB usb hitachi drive. Not sure of the actual disk inside, but probably hitachi.
Last edited by WrinkledCheese; 08-04-2012 at 11:08 AM.
-
May be a re-installation of the operating system in the SSD may help.
-
Re-installation seems like such a Windows solution. I will back up my files and do such. I wish I could have gotten to the bottom of this. I will be especially irritated if the re-installation doesn't resolve the issue...with my computer that is.
-
I have run 125Gb SSD with Ubuntu before and never experienced your problem.
Since you report the problem persists even after the other hard disks have been removed did you get any error log? Your previous error log seems to be related to /dev/sdd1 which should disappear if the hard disks are not there.
-
I did not check for that scenario, I assumed they would.
-
This may be a long shot, but I know that wonky power supplies can cause random, weird problems to a system.
-------------
Folding is Fun
I thought I made a mistake once, but, of course, I was mistaken.
-
I thought of the power supply, and while it is 6 years old it's not a noname brand, although it's not the greatest. It's a Thermaltake.
I'm not sure yet, but so far it seems that I had two bad keyboards. I'm uncertain of how a keyboard could cause a system to lock up, but my wild guess is that some way or another an electric charge is building up on the keybaord - ie static - and then it's freezing on discharge through the USB bus? I dunno sounds a lil far fetched to me but so far I've had a brand new Mac Mini also freezing with one of the keyboards in question and I replaced both keyboards and coincidentally neither system has frozen.
With the Mac Mini the system sorta responded to the keyboard. The mouse would not take it out of sleep, but the press of a key would, but then nothing on the keyboard would make a difference.
I dunno. I guess we'll see tomorrow. If it stops freezing I will update this thread.
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules
|
|