LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - General (http://www.linuxquestions.org/questions/linux-general-1/)
-   -   Remote printing problem becomes total system meltdown - can't even login (http://www.linuxquestions.org/questions/linux-general-1/remote-printing-problem-becomes-total-system-meltdown-cant-even-login-703977/)

jamespetts 02-11-2009 03:57 PM

Remote printing problem becomes total system meltdown - can't even login
 
At home, there are four computers: study, bedroom, kitchen and parlour. Kitchen and parlour are running Ubuntu Intrepid; bedroom is running Ubuntu Hardy and Study is running Windows XP.

Earlier this evening, a problem was found printing from kitchen to parlour. I had earlier configured cups using the GUI to print from kitchen to parlour without difficulty, so found this odd. I spent a great deal of time trying to solve the problem without success, the problem always appearing to be that the server was refusing permission to access it to the clients, even though the server was configured to allow remote printing and remote administration, the clients' IPs were in the allowed list, all relevant usernames were given permission to print, and it had worked with those exact settings recently before.

I tried to configure the printer from bedroom (I had not previously set up the printer from there), and got exactly the same error. I tried to access the CUPS web configuration on parlour, from bedroom and I got a 403 (I had earlier tried the same from kitchen with the same result). So, on bedroom, I tried to configure the printer through Bedroom's CUPS web interface. I added a printer, and selected the appropriate driver (HP Photosmart D7300 series). I sent it a test page from the web interface, but nothing printed - it was marked perpetually as "processing".

I went to parlour to investigate, and could now not access the CUPS web interface even locally - it appeared that the web server was not giving any sort of response. I tried restarting CUPS from the command line, but got a segmentation fault. Trying to reboot from the command line also gave a segmentation fault. I rebooted from the GUI. The loading bar screen appeared, but there were bizarre artefacts (looking like fragments of the blank loading bar under the Ubuntu logo) next to the live loading bar, which progressed as normal.

I reached the login screen as normal, but any attempt to login just brought the interface back to the login screen again, even with "failsafe terminal" selected as the session. I tried rebooting in failsafe mode from the GRUB menu, and attempted to use the automatic repair tools there (FSCK, XFIX, and DPKG fix), but, on rebooting again, there was no change. I am at a total loss, and very concerned that something quite drastic but utterly incomprehensible has gone wrong. I cannot even access shared NFS folders on parlour from other computers. I should be very, very grateful for any assistance at all, especially as this is my parents' computer that I help to maintain for them and they really do rely on it for a great many things.

camorri 02-12-2009 09:10 AM

The symptoms sounds like some kind of hardware failure. You could be experiencing a memory failure, a system board or even a HD failure.

You might want to see if you can find a Knoppix CD, boot it, and run Memtest86 for a period of time. If that is good, then I would look to the HD as most probable cause of failure.

If you have never run memtest, it can be looped for an extended run overnight. It will show you the results on the sceen, I would not run any type of HD diags until you know the system board and ram are O.K. Too much chance of trashing everything on the drive.

You can google memtest86. It is available form a number of sources.

jamespetts 02-12-2009 02:00 PM

Camorri,

Thank you very much for your reply. I have already run Memtest86+ for about 15 hours, with no errors reported. The symptoms do not seem quite consistent with a hard drive failure (the reports of segmentation faults when trying to reboot; I have had a hard drive failure recently in another machine, and the symptoms were much different and the pattern more consistent), although I can see why there might be a suspicion of some hardware failure.

How would I run diagnostics of the motherboard? And would HDParm help with HDD diagnostics?

Thank you again for your reply - it is very much appreciated.

Update: I have tried booting an Ubuntu live CD, and that works fine: I can mount the hard drive, view its contents, and the system appears to work exactly as it normally does when a live CD is used. I cannot find any trace of a hardware failure.

camorri 02-12-2009 06:18 PM

Quote:

How would I run diagnostics of the motherboard?
There may be diagnostics available, depending on the maker of the board. Check the manufactures web site. If memtest ran 15 hours error free, then your system board is probably O.K. If the system will boot a live CD, the system hardware should be good.

Power can also give you some strange problems. Is the system plugged into a properly grounded outlet?

I don't think Hdparm will help much. That is used to change DH parameters.

Right now I would try to run 'fsck' command to check out the file systems on your partitions. If you can, run it from a live CD. Then you can test the root file system. Even if the HD didn't cause the problem i would guess you may find some errors caused by what ever. See the man page, there are lots of options.

Hope this helps.

jamespetts 02-13-2009 02:58 AM

Cliff,

thank you for your help. The computer is indeed plugged into a properly earthed outlet, and has been since 2002. I have already run fsck from the failsafe option in the GRUB menu, and it made no difference to the problem: I still cannot get passed the login screen when booting from the hard drive. Thank you again for your help.

camorri 02-13-2009 05:42 AM

Hmmm, can you log in if you go Ctrl + Alt + F1 ? I must admit, I,m running out of ideas. It may be time for you to re-install the system.

jamespetts 02-13-2009 10:28 AM

Cliff,

thank you for your reply. I have not yet checked CTRL + ALT + F1, but I have found that dmesg produces a few error reports. I am not quite sure what they mean, however: here is a copy -

Code:

[    8.235527] kjournald starting.  Commit interval 5 seconds
[    8.235543] EXT3-fs: mounted filesystem with ordered data mode.
[    8.659624] end_request: I/O error, dev sr0, sector 1431184
[    8.659635] Buffer I/O error on device sr0, logical block 178898
[    9.266496] end_request: I/O error, dev sr0, sector 1431184
[    9.266507] Buffer I/O error on device sr0, logical block 178898
[    9.873350] end_request: I/O error, dev sr0, sector 1431184
[    9.873354] Buffer I/O error on device sr0, logical block 178898
[  10.480213] end_request: I/O error, dev sr0, sector 1431184
[  10.480221] Buffer I/O error on device sr0, logical block 178898
[  11.087082] end_request: I/O error, dev sr0, sector 1431184
[  11.087090] Buffer I/O error on device sr0, logical block 178898
[  11.264216] usb-storage: device scan complete
[  11.264701] scsi 6:0:0:0: Direct-Access    HP      Photosmart D7300 1.00 PQ: 0 ANSI: 2
[  11.274478] sd 6:0:0:0: [sdb] Attached SCSI removable disk
[  11.274674] sd 6:0:0:0: Attached scsi generic sg4 type 0
[  11.518500] kjournald starting.  Commit interval 5 seconds
[  11.518517] EXT3-fs: mounted filesystem with ordered data mode.
[  11.693943] end_request: I/O error, dev sr0, sector 1431184
[  11.693953] Buffer I/O error on device sr0, logical block 178898
[  12.604241] end_request: I/O error, dev sr0, sector 1431184
[  12.604250] Buffer I/O error on device sr0, logical block 178898
[  13.017455] kjournald starting.  Commit interval 5 seconds
[  13.017472] EXT3-fs: mounted filesystem with ordered data mode.
[  13.211187] end_request: I/O error, dev sr0, sector 1431184
[  13.211198] Buffer I/O error on device sr0, logical block 178898
[  14.784902] kjournald starting.  Commit interval 5 seconds
[  14.784918] EXT3-fs: mounted filesystem with ordered data mode.
[  16.924584] ISO 9660 Extensions: Microsoft Joliet Level 3
[  17.069405] ISO 9660 Extensions: RRIP_1991A
[  17.617073] aufs 20080922

Do you think that this might be anything to do with it? ("sr0" is also mentioned earlier in the dmesg output, here:)

Code:

[    5.488463] sr0: scsi3-mmc drive: 4x/40x writer dvd-ram cd/rw xa/form2 cdda tray
[    5.488471] Uniform CD-ROM driver Revision: 3.20
[    5.488607] sr 5:0:0:0: Attached scsi CD-ROM sr0
[    5.491587] sr1: scsi3-mmc drive: 0x/50x cd/rw xa/form2 cdda tray
[    5.491743] sr 5:0:1:0: Attached scsi CD-ROM sr1


camorri 02-13-2009 11:07 AM

I did a little googling on the Buffer I/O error. It seems others have had problems with this as well.

I saw one fix, the user changed changed the cable to the drive from a 40 pin cable to a 80 pin cable. Another fix as to change the BIOS setting for the drive from 'auto' to udma44.

I have no idea if these are the same problems, or not. If the machine has run for a long time, and the cables were not touched, then I would not suspect the cable.

Any chance the BIOS settings have been changed? Some systems have a battery to preserve these settings. If the battery is old, it may start causing settings to change when power is off. Most good boards have the battery clipped in. It is about the size of a quarter. Most are good for about five years. You can buy new ones at any good PC store. If you look at the battery they are identified on one side. So, easy to make sure you get the correct one.

What doesn't make sense, though, if the sr0 settings have changed, why that might affect the logon?

Cut and paste the buffer I/O error into wwww.google.com/linux and have a look. You should be able to find the threads I found.

jamespetts 02-13-2009 05:24 PM

Cliff,

thank you for your replies. The problem is indeed perplexing, especially given that the hard drive is SATA, and appears to work fine when mounted from the Live CD bootup. "sr0" appears from the second quote to refer to the DVD-RAM drive, not the hard drive, but that is the same drive used for the boot from the Live CD, which works, despite the errors. Perhaps I should check the dmesg output for booting into failsafe command line mode without the Live CD? It really does not look like a hardware error, though. And I think that I already have an 80 pin cable going to that DVD-RAM drive.

Edit: The problems to which reference is made on the pages to which I am referred by the Google search that you suggested are all exclusive to CD/DVD drives, which is likely to be irrelevant to my problem.

As to the BIOS issue, I cannot imagine how it could be the BIOS, since the problem first occurred during system operation, not after a reboot (the segmentation faults on trying to restart CUPS and on trying to reboot). It seems that something somewhere has become very seriously corrupted without any obvious cause, or the slightest clue as to where the problem is. Even if I try to reinstall, what is to stop this from happening again and again and again?

camorri 02-13-2009 07:01 PM

I was aware the ones I found with Google were all CD or DVD problems, and sr0 is you DVD.

Have a look at the BIOS settings, and make sure they are as you expect. If they are, and none have changed, I would say you have eliminated that as a possibility. Just keep in mind if anything there has changed, that should not happen. It is just one thing to look at.

Quote:

Even if I try to reinstall, what is to stop this from happening again and again and again?
I have to agree here. You will only know if that is what you decide to do.

Any possibility you experienced a brown out while the machine was being used? Thunderstorm etc?

Another possibility, you could have a power supply starting to fail. If that is the case, the problem will get worse.

jamespetts 02-14-2009 11:45 AM

Cliff,

thank you again for your help. Incidentally, I tried CTRL + ALT + F1. The text only login screen does not work, either: when I type the username and password, it waits about two seconds, and simply presents me with the login prompt again. When I enter an invalid username or password, however, it does correctly state, "login incorrect".

I have checked the BIOS - everything there is as expected. I also doubt that it was a brownout - there were no thunderstorms at the time, and no problems were noticed on the three other computers in the house. How would I diagnose a PSU failure? When the machine was running from the Live CD, there were no noticeable problems, which seems to suggest against a PSU failure.

Thank you very much again for your help.

camorri 02-14-2009 02:19 PM

Quote:

How would I diagnose a PSU failure?
If you have a good meter, you can check the voltages. Other than that, the only other way is to swap it with another, to see if that helps.

This problem is not obvious at all. The fact that you can not log in from the command line either would lead me to believe the files with the users and passwords have been affected. It is possible to reset a user password, with a live CD. Its been a while since I did that. I know there are instructions around on the how-to. It might be worth a try, to reset a user password, to see if you could get logged in and then further diagnose the problems.

I looked up a set of instructions. If you boot the system with a live CD, you need to become root.
Mount the partition where the linux system is. ie 'mount -o rw /dev/hda1 /mnt/hda1' Change the names to match your hardware.

Then 'chroot /mnt/hda1 passwd username2change' you should get prompted to change the password.

If that works, you will at least know one user, and password. Give that a try to log in...

jamespetts 02-14-2009 02:59 PM

Cliff,

hmm, the problem is not so much that it is not recognising the usernames and passwords: when I entered valid usernames and passwords (I tried more than one), it would simply return to the login prompt, but when I entered usernames and passwords that I knew to be invalid, it would correctly give the "login incorrect" error. Thus, it does distinguish between correct username/password pairs and incorrect ones; it is just that, after it has identified the pair as valid, it fails, and returns to the login prompt.

Edit: This does seem to be a login-specific problem...

ozone1410 06-12-2011 06:40 AM

problem
 
i hf similar problem while logging into ubuntu 10.04..
The login screen appears again wheneva i enter the correct loginname and password. bt yes,it does say "Authentication Failure" when i enter invalid password.. it cant be a hardware problem..
i hf no clue what exactly is wrong n i do not want to re-install it.. n as u said thr iz no point formatting n re-installing it and then having d same problem again.. is there any fix around the problem?? If anyone out there got a solution please lemme kno..
Thanx a lot!


All times are GMT -5. The time now is 02:31 PM.