Hi guys,
I'm feeling a little out of my depth here on this one. This box was built two Autumns ago (1998) and has only had a monitor change since. Using RH6.0.
Tonight, on two seperate occassions, between 30 seconds and one minute into my dialup Internet connection, my machine has totally locked up.
The only commonality has been the dialup and corresponding disk activity (sendmail and fetchnews kicking in).
I have had an in-between dialup connection that stayed up fine.
The screen freezes. No mouse, no keyboard. The four consoles do not respond. No beeps occur on keyboard buffer overflow as you'd normally expect. The modem stays on-line, but Rx/Tx lights go dead. The hard disk stops rattling completely (I can normally hear it all the time).
I have 64MB RAM, 128MB swap. gkrellm shows swap to be at or less than 1/4 used, RAM between half and 3/4 used. I run GNOME 1.0.53 and E 0.16 and on both occassions had two Eterms up.
Absolutely nothing out of the normal, bar having 'daytime' uncommented in /etc/services (I was trying out a networking code sample from a book). I have now commented it back out.
I was not accessing audio/cdrom devices, nor floppy (LS-120).
I am using kernel 2.2.12, and have been for smoe time.
The only thing that comes to mind is possibly swapfile corruption, but I don't even know if it's possible with Linux/ext2 or whatever.
Unless the problem is an apparently random hardware fault. I must admit since using GNOME again this past few days memory has been a little lacking.
Ideas?
[jg@cyberstorm jg]$ df Filesystem 1k-blocks Used Available Use% Mounted on /dev/hdb1 2915904 2033150 731959 74% / /dev/hda6 2028098 770226 1153050 40% /home /dev/hda5 2096160 1995712 100448 95% /mnt/data /dev/hda1 2096160 1902592 193568 91% /mnt/windows
[jg@cyberstorm jg]$ free total used free shared buffers cached Mem: 63320 61892 1428 52792 1960 31316 -/+ buffers/cache: 28616 34704 Swap: 128988 20 128968
[jg@cyberstorm jg]$ quota Disk quotas for user jg (uid 500): none
jg@cyberstorm.demon.co.uk wrote:
Hi guys,
I'm feeling a little out of my depth here on this one. This box was built two Autumns ago (1998) and has only had a monitor change since. Using RH6.0.
Tonight, on two seperate occassions, between 30 seconds and one minute into my dialup Internet connection, my machine has totally locked up.
The only commonality has been the dialup and corresponding disk activity (sendmail and fetchnews kicking in).
I have had an in-between dialup connection that stayed up fine.
The screen freezes. No mouse, no keyboard. The four consoles do not respond. No beeps occur on keyboard buffer overflow as you'd normally expect. The modem stays on-line, but Rx/Tx lights go dead. The hard disk stops rattling completely (I can normally hear it all the time).
I have 64MB RAM, 128MB swap. gkrellm shows swap to be at or less than 1/4 used, RAM between half and 3/4 used. I run GNOME 1.0.53 and E 0.16 and on both occassions had two Eterms up.
Absolutely nothing out of the normal, bar having 'daytime' uncommented in /etc/services (I was trying out a networking code sample from a book). I have now commented it back out.
I was not accessing audio/cdrom devices, nor floppy (LS-120).
I am using kernel 2.2.12, and have been for smoe time.
The only thing that comes to mind is possibly swapfile corruption, but I don't even know if it's possible with Linux/ext2 or whatever.
Unless the problem is an apparently random hardware fault. I must admit since using GNOME again this past few days memory has been a little lacking.
Ideas?
[jg@cyberstorm jg]$ df Filesystem 1k-blocks Used Available Use% Mounted on /dev/hdb1 2915904 2033150 731959 74% / /dev/hda6 2028098 770226 1153050 40% /home /dev/hda5 2096160 1995712 100448 95% /mnt/data /dev/hda1 2096160 1902592 193568 91% /mnt/windows
[jg@cyberstorm jg]$ free total used free shared buffers cached Mem: 63320 61892 1428 52792 1960 31316 -/+ buffers/cache: 28616 34704 Swap: 128988 20 128968
I don't like to be a pessimist, but that sort of thing sounds like a mobo problem. Possibly the power regulation on the board is failing, may be a duff decoupling capacitor. If it is the mobo, the is no real way to test it short of wapping it out and seeing whether that cures the problem.
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
On Tue, Feb 29, 2000 at 09:39:37AM +0000, raph wrote:
jg@cyberstorm.demon.co.uk wrote: I don't like to be a pessimist, but that sort of thing sounds like a mobo problem. Possibly the power regulation on the board is failing, may be a duff decoupling capacitor. If it is the mobo, the is no real way to test it short of wapping it out and seeing whether that cures the problem.
Hmmm - but surely if he wapps the mobo out then nowts gunna work ;)
Brett (in an ever helpful way)
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
Brett Parker wrote:
On Tue, Feb 29, 2000 at 09:39:37AM +0000, raph wrote:
jg@cyberstorm.demon.co.uk wrote: I don't like to be a pessimist, but that sort of thing sounds like a mobo problem. Possibly the power regulation on the board is failing, may be a duff decoupling capacitor. If it is the mobo, the is no real way to test it short of wapping it out and seeing whether that cures the problem.
Hmmm - but surely if he wapps the mobo out then nowts gunna work ;)
But then 'swap' implies an exchange. Mind you, if he exchanges his mobo for you brain .... I'm not sure who would be the gainer ;-)
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
I have to say, I think motherboard failure is unlikely in this instance as the problem only appears to manifest itself when James is online. If the motherboard was failing, I would expect this to happen at random intervals regardless of the state of the connection.
The only other vague possibility is that the serial controller is asserting a HLT on the processor, but I'd have thought that's unlikely.
Paul
On Tue, Feb 29, 2000 at 10:05:19AM +0000, raph wrote:
Brett Parker wrote:
On Tue, Feb 29, 2000 at 09:39:37AM +0000, raph wrote:
jg@cyberstorm.demon.co.uk wrote: I don't like to be a pessimist, but that sort of thing sounds like a mobo problem. Possibly the power regulation on the board is failing, may be a duff decoupling capacitor. If it is the mobo, the is no real way to test it short of wapping it out and seeing whether that cures the problem.
Hmmm - but surely if he wapps the mobo out then nowts gunna work ;)
But then 'swap' implies an exchange. Mind you, if he exchanges his mobo for you brain .... I'm not sure who would be the gainer ;-)
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
On Tue, 29 Feb 2000, raph wrote:
I don't like to be a pessimist, but that sort of thing sounds like a mobo problem.
A what?
James.
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
jg@cyberstorm.demon.co.uk wrote:
Hi guys,
I'm feeling a little out of my depth here on this one. This box was built two Autumns ago (1998) and has only had a monitor change since. Using RH6.0.
Tonight, on two seperate occassions, between 30 seconds and one minute into my dialup Internet connection, my machine has totally locked up.
The only commonality has been the dialup and corresponding disk activity (sendmail and fetchnews kicking in).
I have had an in-between dialup connection that stayed up fine.
The screen freezes. No mouse, no keyboard. The four consoles do not respond. No beeps occur on keyboard buffer overflow as you'd normally expect. The modem stays on-line, but Rx/Tx lights go dead. The hard disk stops rattling completely (I can normally hear it all the time).
I have 64MB RAM, 128MB swap. gkrellm shows swap to be at or less than 1/4 used, RAM between half and 3/4 used. I run GNOME 1.0.53 and E 0.16 and on both occassions had two Eterms up.
Absolutely nothing out of the normal, bar having 'daytime' uncommented in /etc/services (I was trying out a networking code sample from a book). I have now commented it back out.
I was not accessing audio/cdrom devices, nor floppy (LS-120).
I am using kernel 2.2.12, and have been for smoe time.
the best thing I can suggest is to upgrade to the latest stable kernel and remake your swap file if you think it is corrupt, I have had a corrupt swap file before and it does do weird things! One more thing... UPDATES !!!.. there have been a lot of security updates since 6.0 was released, I'm not sure if any of these allow remote lockups?? but if your not sure then upgrade all the necessary packages in the redhat updates directory on your local mirror...
this is one thing I will shout from the highest rooftop to to everybody out there who has machines which are online in anyway, updates updates and updates;)..
give these a go and see what happens
Scoobz
The only thing that comes to mind is possibly swapfile corruption, but I don't even know if it's possible with Linux/ext2 or whatever.
Unless the problem is an apparently random hardware fault. I must admit since using GNOME again this past few days memory has been a little lacking.
Ideas?
[jg@cyberstorm jg]$ df Filesystem 1k-blocks Used Available Use% Mounted on /dev/hdb1 2915904 2033150 731959 74% / /dev/hda6 2028098 770226 1153050 40% /home /dev/hda5 2096160 1995712 100448 95% /mnt/data /dev/hda1 2096160 1902592 193568 91% /mnt/windows
[jg@cyberstorm jg]$ free total used free shared buffers cached Mem: 63320 61892 1428 52792 1960 31316 -/+ buffers/cache: 28616 34704 Swap: 128988 20 128968
[jg@cyberstorm jg]$ quota Disk quotas for user jg (uid 500): none
-- James Green jg@cyberstorm.demon.co.uk Site Manager jg@linuxnewbie.com LinuxNewbie.com LNC http://www.cyberstorm.demon.co.uk/
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
On Tue, 29 Feb 2000, Neill Newman wrote:
the best thing I can suggest is to upgrade to the latest stable kernel
Upgrading tonight.
and remake your swap file if you think it is corrupt, I have had a
Um, anyone got the cmd to do this? And is it safe within console or do I need to be on a floppy disk or something?
corrupt swap file before and it does do weird things!
;)
One more thing... UPDATES !!!.. there have been a lot of security updates since 6.0 was released, I'm not sure if any of these allow remote lockups?? but if your not sure then upgrade all the necessary packages in the redhat updates directory on your local mirror...
Well, I have a 6.1 cd.
Question: how does one perform an rpm -Uvh * within the rpms directory and *only* install package which require upgrading, i.e. I don't want to install all possible packages.
James.
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
Green J M K wrote:
On Tue, 29 Feb 2000, Neill Newman wrote:
the best thing I can suggest is to upgrade to the latest stable kernel
Upgrading tonight.
and remake your swap file if you think it is corrupt, I have had a
Um, anyone got the cmd to do this? And is it safe within console or do I need to be on a floppy disk or something?
man mkswp ;)..
Sz
Hi. I've got a broken Debian package: dpkg --purge giram-mesa (Reading database ... 80208 files and directories currently installed.) Removing giram-mesa ... /usr/share/doc/giram-mesa/Tutorial: No such file or directory dpkg: error processing giram-mesa (--purge): subprocess pre-removal script returned error exit status 1 Errors were encountered while processing: giram-mesa So, what's the "nuke the *&@$" option for dpkg??? Bill ------------------------------------------------------------------------- W.B.Hill \ Email: wbh@uea.ac.uk \ 492 Earlham Road \ Home: (+44) 1603 455069 \ "Nothing is idiot-proof Norwich NR4 7HP. \ Fax: 0870 054 7508 \ given the right idiot" United Kingdom \ Mobile: 0410 781584 \ (+44) 1603 255141 \ Web: http://wbh.org \ #include <stdisclaimer.h> ------------------------------------------------------------------------- PGP-Fingerprint: B3 6F 2E D6 - B6 48 F1 FB -*- 7C AA 28 0C - 96 61 9A B5
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
On Tue, Feb 29, 2000 at 07:42:01PM +0000, wbh wrote:
Hi. I've got a broken Debian package: dpkg --purge giram-mesa (Reading database ... 80208 files and directories currently installed.) Removing giram-mesa ... /usr/share/doc/giram-mesa/Tutorial: No such file or directory dpkg: error processing giram-mesa (--purge): subprocess pre-removal script returned error exit status 1 Errors were encountered while processing: giram-mesa So, what's the "nuke the *&@$" option for dpkg???
Your best bet is to edit /var/lib/spkg/info/giram-mesa.prerm and take out a -e in the first line (or remove a "set -e" in the script somewhere), which will stop it stopping on errors, meaning that it will merrily plough on even if the world ends :-)
Alternatively, touching /usr/share/doc/giram-mesa/Tutorial might be safer?
Aquarius
Your best bet is to edit /var/lib/spkg/info/giram-mesa.prerm and take out a -e in the first line (or remove a "set -e" in the script somewhere), which will stop it stopping on errors, meaning that it will merrily plough on even if the world ends :-)
Alternatively, touching /usr/share/doc/giram-mesa/Tutorial might be safer?
That got it - thanks! Bill ------------------------------------------------------------------------- W.B.Hill \ Email: wbh@uea.ac.uk \ 492 Earlham Road \ Home: (+44) 1603 455069 \ "Nothing is idiot-proof Norwich NR4 7HP. \ Fax: 0870 054 7508 \ given the right idiot" United Kingdom \ Mobile: 0410 781584 \ (+44) 1603 255141 \ Web: http://wbh.org \ #include <stdisclaimer.h> ------------------------------------------------------------------------- PGP-Fingerprint: B3 6F 2E D6 - B6 48 F1 FB -*- 7C AA 28 0C - 96 61 9A B5
[ This email came to you via the Anglian Linux User Group list ] [ If you only wish to recieve event announcements, email the ] [ SUBJECTs of "unsubscribe alug" and "subscribe alug-announce" ] [ to listserver@stu.uea.ac.uk -- We do need your support, tho' ]
-
Aquarius -
Brett Parker -
Green J M K -
jg＠cyberstorm.demon.co.uk -
Neill Newman -
Paul Russell -
raph -
wbh