Author Topic: MD PXE Boot, used to work, now fails  (Read 1645 times)

dison4linux

  • Regular Poster
  • **
  • Posts: 19
    • View Profile
MD PXE Boot, used to work, now fails
« on: February 22, 2010, 09:34:11 pm »
Getting message:
/init: .: line 1: can't open /temp/net-eth0.conf
kernel panic - syncing: attempted to kill init!

When booting Media Director

After searching the forums, I followed the directions on the wiki for adding module that are at http://wiki.linuxmce.org/index.php/Unrecognized_NIC#Modify_modules_on_core

No change.  By the way my ethernet device uses module e100.

I have also tried removing the device from the media directors page in web admin, and re-running the diskless build scripts.  I have also re-booted the core several times.

Everything was working fine before, and then I had to turn off the media director to move the power cord and after I booted it back up I got these errors.

wierdbeard65

  • Guru
  • ****
  • Posts: 449
    • View Profile
    • My Quest
Re: MD PXE Boot, used to work, now fails
« Reply #1 on: February 22, 2010, 09:40:39 pm »
I too am having this problem. See thread http://forum.linuxmce.org/index.php?topic=9656.0

So far, no luck  :(

I really hope we can find a solution to this one!
Paul
If you have the time to help, please see where I have got to at: http://wiki.linuxmce.org/index.php/User:Wierdbeard65

purps

  • NEEDS to work for LinuxMCE
  • ***
  • Posts: 1374
  • If it ain't broke, tweak it
    • View Profile
Re: MD PXE Boot, used to work, now fails
« Reply #2 on: February 22, 2010, 09:57:01 pm »
dison, I had these problems, and you know what it turned out to be? Dead stick of RAM. If you really haven't changed anything, software or hardware, then give your MD hardware a complete check.
1004 RC :: looking good :: upgraded 01/04/2013
my setup :: http://wiki.linuxmce.org/index.php/User:Purps

dison4linux

  • Regular Poster
  • **
  • Posts: 19
    • View Profile
Re: MD PXE Boot, used to work, now fails
« Reply #3 on: February 23, 2010, 02:06:06 am »
I took purps' advice and tried PXE booting a whole different computer.  It did still use the e100 module however.  Similar results.
IP-Config: eth0 hardware address 00:09:6b:38:c4:5c mtu 1500 DHCP RARP
e100: eth0: e100_watchdog: link up, 100Mbps, full-duplex

That is right before where it would stop before and say
/init: .: line 1: can't open /temp/net-eth0.conf
kernel panic - syncing: attempted to kill init!


Any other ideas?

I suppose I could try to PXE boot a laptop or something that uses a different module...?

wierdbeard65

  • Guru
  • ****
  • Posts: 449
    • View Profile
    • My Quest
Re: MD PXE Boot, used to work, now fails
« Reply #4 on: February 23, 2010, 02:15:38 pm »
Unfortunately, I don't have any other machines to try, so will have to let you do this one!

It would seem to me that perhaps something has been broken in the e100 module for the pxe boot? We are both experiencing exactly the same error here, so I don't think it can be a hardware problem (at least not as in faulty hardware).

If anyone knows how to rebuild the pxe kernel, I can give that a shot, but otherwise....
Paul
If you have the time to help, please see where I have got to at: http://wiki.linuxmce.org/index.php/User:Wierdbeard65

purps

  • NEEDS to work for LinuxMCE
  • ***
  • Posts: 1374
  • If it ain't broke, tweak it
    • View Profile
Re: MD PXE Boot, used to work, now fails
« Reply #5 on: February 23, 2010, 02:38:14 pm »
I took purps' advice and tried PXE booting a whole different computer.  It did still use the e100 module however.  Similar results.
IP-Config: eth0 hardware address 00:09:6b:38:c4:5c mtu 1500 DHCP RARP
e100: eth0: e100_watchdog: link up, 100Mbps, full-duplex

That is right before where it would stop before and say
/init: .: line 1: can't open /temp/net-eth0.conf
kernel panic - syncing: attempted to kill init!


Any other ideas?

I suppose I could try to PXE boot a laptop or something that uses a different module...?

I'm not sure that's conclusive. If you said it was the SAME problem, then fine, but you said the problems were SIMLIAR. Have you run memtest? Have you tried unplugging absolutely everything, except the network cable? If you have more than one stick of RAM, try unplugging one at a time.

You could make a PXE boot disk and see if that gets it going. http://rom-o-matic.net/ I used this to netboot an ancient machine with no PXE ROM chip whatsoever! I used the etherboot one btw, not the gPXE one.
1004 RC :: looking good :: upgraded 01/04/2013
my setup :: http://wiki.linuxmce.org/index.php/User:Purps

wierdbeard65

  • Guru
  • ****
  • Posts: 449
    • View Profile
    • My Quest
Re: MD PXE Boot, used to work, now fails
« Reply #6 on: February 23, 2010, 03:26:54 pm »
Hi Purps!

I honestly think that a hardward fault is not to blame here.

I am expreiencing the same error and the only thing in common (from the thread) is e100. I can also boot my MD from a live CD and it works fine, so I have no reason to suspect faulty hardware.

I don't understand how the pxe files are packaged up etc. My suspicion is that the kernel that is loaded as part of the pxe process has been recently rebuilt and either e100 support has been removed, or is faulty. If someone can point me at a resource that explains how to re-compile the kernel that is delivered via PXE, then I camn (possibly) re-compile it (with e100!) and test my theory.

Alternatively, if someone more experiance than I can do this.... Who builds ther kernels used by MCE? Can they confirm if any changes to the e100 support have been made?
Paul
If you have the time to help, please see where I have got to at: http://wiki.linuxmce.org/index.php/User:Wierdbeard65

dison4linux

  • Regular Poster
  • **
  • Posts: 19
    • View Profile
Re: MD PXE Boot, used to work, now fails
« Reply #7 on: February 23, 2010, 06:55:31 pm »
Yeah it was exactly the same error.  There is a 60 second timeout before the kernel panic.  I didn't wait the 60 seconds before I started my last post.  I am with wierdbeard65, I do not think it was the hardware, as I tried a whole different computer, with different memory.  The only software change I would have done is to update sqlcvs.  Once I get all my DVD ISOs backed up onto another computer, I will probably format/re-install unless anyone has any other ideas.

wierdbeard65

  • Guru
  • ****
  • Posts: 449
    • View Profile
    • My Quest
Re: MD PXE Boot, used to work, now fails
« Reply #8 on: February 24, 2010, 10:56:22 pm »
Any luck with this? I don't want to have to chuck out my box, but need to get an MD running soon :)
Paul
If you have the time to help, please see where I have got to at: http://wiki.linuxmce.org/index.php/User:Wierdbeard65

dison4linux

  • Regular Poster
  • **
  • Posts: 19
    • View Profile
Re: MD PXE Boot, used to work, now fails
« Reply #9 on: February 25, 2010, 06:47:51 pm »
No, I cannot PXE boot any computers off of the core.  Something must have gotten screwed up possibly when I updated sqlCVS or maybe during an apt-get update.  Either way I am going to wait until another beta or RC comes out (I understand it may be months) and format and re-install.

huh

  • Guru
  • ****
  • Posts: 235
    • View Profile
Re: MD PXE Boot, used to work, now fails
« Reply #10 on: February 25, 2010, 07:11:35 pm »
I'm confused with the symptoms those posting here are experiencing.  One of the post below seems to say that the MD will boot but the PXE image is not found so the MD PXE times out.

I have a couple revos that I used to be able to use WOL but it no longer works even though I didn't change anything.  I have tried etherwake and have tried waking from an orbiter.   There is no confirmation (and never to my knowledge has been) that the message has been sent like there is when you shut down the MD.

I wasn't going to post until I tried from a straight ubuntu system to see if something was wrong on my network but saw the last post from dison4linux and it looks like we have the same symptoms.  If I were to guess, this hasn't worked since last Nov or so.


dison4linux

  • Regular Poster
  • **
  • Posts: 19
    • View Profile
Re: MD PXE Boot, used to work, now fails
« Reply #11 on: March 02, 2010, 03:24:46 am »
After running apt-get update, again, to get: Update 2010-02-28 22812 installed.

I deleted the media director from the web interface, ran Diskless_BuildDefaultImage.sh & Diskless_CreateTBZ.sh again and now everything is working as before.  Same e100 module, same computer with the "suspect" RAM.  The only thing that seemed to change was the recent apt-get update.

Go figure.

purps

  • NEEDS to work for LinuxMCE
  • ***
  • Posts: 1374
  • If it ain't broke, tweak it
    • View Profile
Re: MD PXE Boot, used to work, now fails
« Reply #12 on: March 03, 2010, 10:02:42 pm »
Glad it's working again mate. Will just have to keep an eye on it.

FYI I got a replacement RAM kit, got the same problem as before..... turned out to be ANOTHER bad stick! On my "third time lucky" kit now, and seems to be working fine *touches wood*
1004 RC :: looking good :: upgraded 01/04/2013
my setup :: http://wiki.linuxmce.org/index.php/User:Purps