Client Comp: Failed to mount root filesystem!

2008-06-24
2014-03-12
  • Jason Penedo

    Jason Penedo - 2008-06-24

    I have been working on this for two straight days now and really am lost on what to do. I have been reading this forum and a few others for help but to no avail. I am trying to set up a server with Ubuntu and using clonezilla to be able to image computers and put the image on others. I have followed step by step this site:

    http://packratstudios.com/index.php/2008/04/20/how-to-setup-clonezilla-on-linux-ubuntu-quick-start-guide/

    However, it did not work for me. I then tried this site (it wasn't much different but did have its minor differances):

    https://wiki.edubuntu.org/SettingUpClonezillaDRBLonUbuntu

    And I am still running into the same problems. The error I am getting is when I boot from a client, it starts looking for a DHCP server. It does not get it on the first attempt but does on its second (I guess because I am on a large network). That is not the problem though. Once it gets the IP, it loads a few lines then I quickly see the clonezilla white background screen. It then comes right off and loads a ton more of lines. Most are gibberish and I cannot make out. It eventually comes up to an error and the following is the only things I see on the last screen:

    Mounting root filesystem 10.1.11.51:/tftpboot/node_root at /sysrool with option "gibberish"
    mount: RPC: Remote system error - network is unreachable
    Failed to mount the root filesystem! Something went wrong!!!
    Only IP address offered by this DHCP server name will be accepted: drbl
    DHCP server name is drbl: 
    Successfully obtained IP address via wired link [eth0]
    **************************************************************
    FATAL ERROR: Failed to mount root filesystem!!! Please check:
    1) Driver of network card loaded successfully?
    2) TCPwrapper setting
    3) Any other DHCP server?
    4) DHCP server is able to reverse-map
    **************************************************************
    You can enter the shell if you want to debug...
    Kernel panic - not syncing: Attempted to kill init!

    I went through all the suggestions and can't figure it out. I will provide any information that is requested here. I really would appreciate any help whatsoever. Thanks!

    BTW, in /var/log/daemon.log I found these. This is just a small snippet of a long page. It is very long and I am guessing tons of computers are going to it by mistake. Here is the only part that relates to the MAC address that is suppose to be going to it and not working (the machine I am getting the error on):

    Jun 24 13:01:27 ntsrvpdclone dhcpd: DHCPDISCOVER from 00:80:45:47:a0:da via eth0
    Jun 24 13:01:28 ntsrvpdclone dhcpd: DHCPOFFER on 10.1.0.11 to 00:80:45:47:a0:da via eth0
    Jun 24 13:01:30 ntsrvpdclone dhcpd: DHCPREQUEST for 10.1.0.11 (10.1.10.39) from 00:80:45:47:a0:da via eth0
    Jun 24 13:01:30 ntsrvpdclone dhcpd: DHCPACK on 10.1.0.11 to 00:80:45:47:a0:da via eth0
    Jun 24 13:01:30 ntsrvpdclone in.tftpd[14844]: tftp: client does not accept options 
    Jun 24 13:01:40 ntsrvpdclone dhcpd: DHCPDISCOVER from 00:80:45:47:a0:da via eth0
    Jun 24 13:01:40 ntsrvpdclone dhcpd: DHCPOFFER on 10.1.0.11 to 00:80:45:47:a0:da via eth0
    Jun 24 13:01:41 ntsrvpdclone dhcpd: DHCPREQUEST for 10.1.15.66 (10.1.2.25) from 00:80:45:47:a0:da via eth0: unknown lease 10.1.15.66.
    Jun 24 13:01:41 ntsrvpdclone dhcpd: DHCPDISCOVER from 00:80:45:47:a0:da via eth0
    Jun 24 13:01:41 ntsrvpdclone dhcpd: DHCPOFFER on 10.1.0.11 to 00:80:45:47:a0:da via eth0
    Jun 24 13:01:41 ntsrvpdclone dhcpd: DHCPREQUEST for 10.1.0.11 (10.1.10.39) from 00:80:45:47:a0:da via eth0
    Jun 24 13:01:41 ntsrvpdclone dhcpd: DHCPACK on 10.1.0.11 to 00:80:45:47:a0:da via eth0

    Is this where my problem is? How can I solve it? Thanks!

     
    • Steven Shiau

      Steven Shiau - 2008-06-25

      1. Weird message:
      Mounting root filesystem 10.1.11.51:/tftpboot/node_root at /sysrool with option "gibberish" 
         No idea where this comes from. Did you have 2nd DHCP server in the same network environment ?

      2. Please follow this:
      http://drbl.sourceforge.net/one4all/#bug-report
         to run the command /opt/drbl/bin/drbl-bug-report then post the generated file or mail to me <steven _at_ nchc org tw>

       
      • Jason Penedo

        Jason Penedo - 2008-06-25

        Here is the error. I am going to go ahead and follow the first guide again but I expect the same exact results.

        -------CUT BEGIN HERE--------------------------------------------
        Description of problem:
            Client computer comes up with the error message that I posted
        How reproducible:
            Always
        Steps to Reproduce:
            I followed the guide posted
        Actual results:
            The error
        Expected results:
            Clonezilla to start up and allow me to image or force an image on

        Some info about the DRBL environment (PLEASE DO NOT EDIT THEM!):

        OS version: Ubuntu 8.04
        Server arch: i686
        Server CPU:  Intel(R) Pentium(R) 4 CPU 3.00GHz
        Server memory size: 514764 kB
        Server Kernel version: 2.6.24-19-generic
        Installed DRBL-related packages:  drbl-1.8.2-63 partclone-0.0.6-3 mkpxeinitrd-net-1.2-22 clonezilla-2.2.4-54 mkswap-uuid-0.1.0-3 drbl-partimage-0.6.7-1drbl drbl-ntfsprogs-2.0.0-1 drbl-chntpw-0.0.20040818-7 drbl-lzop-1.02-0.8drbl udpcast-20071228 drbl-etherboot-5.4.3-2 freedos-1.0-11drbl
        Client kernel version: 2.6.24-19-386
        Client kernel arch: i486
        NICs with private IP address in server: eth0 eth0:0
        Private IP address in server: 10.1.10.39 10.1.11.51
        Total client no: 12
        Client IP address: 10.1.0.1 10.1.0.10 10.1.0.11 10.1.0.12 10.1.0.2 10.1.0.3 10.1.0.4 10.1.0.5 10.1.0.6 10.1.0.7 10.1.0.8 10.1.0.9
        ===
        -------CUT END HERE----------------------------------------------

         
    • Jason Penedo

      Jason Penedo - 2008-06-25

      Alright, I got it working when I installed two nics on it. Now I have a few more questions.

      First, everytime a client computer loads up and starts looking for a DHCP server (keeps going DHCP...) it takes forever. Always, after the first attempt, it fails. I have to press ENTER for it to look again and it usually finds it almost immediately. What I found very weird is that I even divided the DRLC part into its own private network with just the server (off the 2nd nic) and 1 client. It still required itself to go to the 2nd try even though the network is only comprised of them. Any ideas what is going on? Any way to fix?

      Another problem/question I have is about the MAC addresses. I did a lookup and only found my client following the instructions. I then confirmed and closed the lookup and it put it into the text file. My question is how can I make this easier. Everytime I want to force an image or get an image off a client, do I have to input their MAC address into that text file AND rerun the setup? I think that really would blow and almost defeats the purpose. Please provide suggestions for me to hurdle this.

      This leads to another question. On the step by step I followed, it suggested to accept the option of just having one IP address for clients. I believed it said that this will still work with multiple clients. Is that true? If I put up 20 clients, would it all be okay even though I set it up for one client? Should I rerun it?

      Finally, I was just informed that we possibly want to put the images on another computer then the server. Is there anyway I can tell the server to store the image files onto this other server? I remember during setup you had to set up a directory (default was /clonezilla/ I believe). Should I just direct it to the other server that we want to store it on? How should I go about doing that?

      I really appreciate all the help everyone. So happy I finally got it!

       
      • Steven Shiau

        Steven Shiau - 2008-06-26

        > First, everytime a client computer loads up and starts looking for a DHCP server (keeps going DHCP...) it takes forever. Always, after the first attempt, it fails. I have to press ENTER for it to look again and it usually finds it almost immediately. What I found very weird is that I even divided the DRLC part into its own private network with just the server (off the 2nd nic) and 1 client. It still required itself to go to the 2nd try even though the network is only comprised of them. Any ideas what is going on? Any way to fix?

        Could you please explain more ? Not so sure what your question is...

        > Another problem/question I have is about the MAC addresses. I did a lookup and only found my client following the instructions. I then confirmed and closed the lookup and it put it into the text file. My question is how can I make this easier. Everytime I want to force an image or get an image off a client, do I have to input their MAC address into that text file AND rerun the setup? I think that really would blow and almost defeats the purpose. Please provide suggestions for me to hurdle this.

        When you run drblpush, you have a chance to collect the MAC addresses of your clients, then when you run "/opt/drbl/sbin/dcs", you can choose which clients to be restored by a MAC addresses list.

        > This leads to another question. On the step by step I followed, it suggested to accept the option of just having one IP address for clients. I believed it said that this will still work with multiple clients. Is that true? If I put up 20 clients, would it all be okay even though I set it up for one client? Should I rerun it?

        Yes. It will still work with multiple clients.

        > Finally, I was just informed that we possibly want to put the images on another computer then the server. Is there anyway I can tell the server to store the image files onto this other server? I remember during setup you had to set up a directory (default was /clonezilla/ I believe). Should I just direct it to the other server that we want to store it on? How should I go about doing that?

        In DRBL server, kernel NFS service is used, then only local file system can be exported. If you want to mount remote dir via network then export it for DRBL clients, you have to use user space mode NFS.

         
        • Jason Penedo

          Jason Penedo - 2008-06-26

          >Could you please explain more ? Not so sure what your question is...

          Everytime a client goes on, I tell it to boot from LAN first. When the computer starts up, it says "DHCP..." and starts trying to gain an IP address from the DRBL server. No matter which computer I am using, it always fails the first time saying it cannot find an address. I always have to click ENTER for it to try again and almost immediately it works (gains the IP address and starts running clonezilla commands). This is extremely annoying and I have no idea for the cause of it. I first thought it was because of the huge network I am on but I set up my own private network with the server working and it still would do the same thing (failing first time, getting it the second time). My question is what is the cause of this? Is there any way to fix?

          >When you run drblpush, you have a chance to collect the MAC addresses of your clients, then when you run "/opt/drbl/sbin/dcs", you can choose which clients to be restored by a MAC addresses list.

          I would like to not have to do that. I will be finishing up this project and it will require someone else to work it. I read that there is a way to simply make it so any computer that boots Network first through PXE could automatically be allowed in. I even have seen you explain how to get it done at other places in this forum. Here is a link to the faq that addresses it (3):

          http://drbl.sourceforge.net/faq/index.php#path=./2_System&entry=37_coexist_with_existing_dhcp.faq

          I read there and from what you said that I have to get to the "/etc/dhcp3/dhcpd.conf" and then uncomment the 'allow members of "DRBL-Client". My problem is I cannot find that anywhere. I do a ctrl-f and search allow members but it is listed only 3 times (in the instructions how to do it within the conf file). Nowhere else is it. Am I suppose to just add the line anywhere without the # sign? If not, where is the line that I uncomment? Is this not going to end up doing what I want?

          >Yes. It will still work with multiple clients.

          I just would like to make sure that I understand you correctly. You are saying that if I set it up for one client address (give it one IP address to use), I can have 20 clients log in at once and it will be okay with just the one addy I gave it? Please confirm so I can move on.

          >In DRBL server, kernel NFS service is used, then only local file system can be exported. If you want to mount remote dir via network then export it for DRBL clients, you have to use user space mode NFS.

          I am as beginner as it gets and I really have no clue what all the means. Just to make sure I am clear, I would like it that when the client goes on and gets imaged, the image file will not go to the clonezilla server but be sent to another server that I have on the network. I verified that the clonezilla can see the server and the folder I want to put it in. How do I make it so all the images are stored there and then when I would like an image to have forced onto another client, it will get them from there.

          Please answer all my questions to the best of your ability. I am wasting this whole week of work on this project and really would like it finished. Try and be as thorough as possible. I know it is a pain but trust me when I say that I really appreciate it!

           
          • Steven Shiau

            Steven Shiau - 2008-06-26

            I already answered you in drbl-user mailing list.

             
            • Jason Penedo

              Jason Penedo - 2008-06-26

              What do you mean? I just wrote these questions and can't find your answers anywhere. Link me please.

               
              • Steven Shiau

                Steven Shiau - 2008-06-26

                My bad...
                Here you are:
                SourceForge.net wrote:
                > > Read and respond to this message at:
                > > https://sourceforge.net/forum/message.php?msg_id=5055678
                > > By: potatoman123
                > >
                > >  
                >> >> Could you please explain more ? Not so sure what your question is...
                >> >>    
                > >
                > > Everytime a client goes on, I tell it to boot from LAN first. When the computer
                > > starts up, it says "DHCP..." and starts trying to gain an IP address from the
                > > DRBL server. No matter which computer I am using, it always fails the first
                > > time saying it cannot find an address. I always have to click ENTER for it to
                > > try again and almost immediately it works (gains the IP address and starts running
                > > clonezilla commands). This is extremely annoying and I have no idea for the
                > > cause of it. I first thought it was because of the huge network I am on but
                > > I set up my own private network with the server working and it still would do
                > > the same thing (failing first time, getting it the second time). My question
                > > is what is the cause of this? Is there any way to fix?
                > >
                > >  
                Maybe the problem is in the network switch. Is that a managed switch ?
                Did you try to use cross-over cable to link your server and client ?
                >> >> When you run drblpush, you have a chance to collect the MAC addresses of your
                >> >>    
                > > clients, then when you run "/opt/drbl/sbin/dcs", you can choose which clients
                > > to be restored by a MAC addresses list.
                > >
                > > I would like to not have to do that. I will be finishing up this project and
                > > it will require someone else to work it. I read that there is a way to simply
                > > make it so any computer that boots Network first through PXE could automatically
                > > be allowed in. I even have seen you explain how to get it done at other places
                > > in this forum. Here is a link to the faq that addresses it (3):
                > >
                > > http://drbl.sourceforge.net/faq/index.php#path=./2_System&entry=37_coexist_with_
                > > existing_dhcp.faq
                > >
                > > I read there and from what you said that I have to get to the "/etc/dhcp3/dhcpd.conf"
                > > and then uncomment the 'allow members of "DRBL-Client". My problem is I cannot
                > > find that anywhere. I do a ctrl-f and search allow members but it is listed
                > > only 3 times (in the instructions how to do it within the conf file). Nowhere
                > > else is it. Am I suppose to just add the line anywhere without the # sign? If
                > > not, where is the line that I uncomment? Is this not going to end up doing what
                > > I want?
                > > 
                > >  
                Just uncomment all of them. That means all of the pools of DHCP clients
                must be the members of DRBL-client then they can join DRBL environment.
                >> >> Yes. It will still work with multiple clients.
                >> >>    
                > >
                > > I just would like to make sure that I understand you correctly. You are saying
                > > that if I set it up for one client address (give it one IP address to use),
                > > I can have 20 clients log in at once and it will be okay with just the one addy
                > > I gave it? Please confirm so I can move on.
                > > 
                > >  
                Oh, I think I misunderstood your question. No, you have to assign 20 IP
                addresses for 20 clients to run simultaneously. You can not assign only
                one IP address when you run drblpush. Give it enough number.
                >> >> In DRBL server, kernel NFS service is used, then only local file system can
                >> >>    
                > > be exported. If you want to mount remote dir via network then export it for
                > > DRBL clients, you have to use user space mode NFS.
                > >
                > > I am as beginner as it gets and I really have no clue what all the means. Just
                > > to make sure I am clear, I would like it that when the client goes on and gets
                > > imaged, the image file will not go to the clonezilla server but be sent to another
                > > server that I have on the network. I verified that the clonezilla can see the
                > > server and the folder I want to put it in. How do I make it so all the images
                > > are stored there and then when I would like an image to have forced onto another
                > > client, it will get them from there.
                > >  
                This is really an advanced job. All the drbl clients share
                /home/partimag via NFS, and the NFS server exports it's /home/partimag.
                Basically for kernel space nfs (This is what DRBL using) won't allow you
                to export a remote resource (like mounted from NFS, samba, sshfs...).
                You can switch the kernel space NFS to userd mode nfs by installing the
                program unfs3 (User-space NFSv3 Server) to replace the original NFS
                program. Another approach is you can try iSCSI or AoE, the low level
                protocol device.

                > > Please answer all my questions to the best of your ability. I am wasting this
                > > whole week of work on this project and really would like it finished. Try and
                > > be as thorough as possible. I know it is a pain but trust me when I say that
                > > I really appreciate it!
                > >
                > >  
                Do my best.
                > > ______________________________________________________________________
                > > You are receiving this email because you elected to monitor this forum.
                > > To stop monitoring this forum, login to SourceForge.net and visit:
                > > https://sourceforge.net/forum/unmonitor.php?forum_id=799287
                > >  

                 
                • Jason Penedo

                  Jason Penedo - 2008-06-26

                  Hmm, well during drlpush, I said no to configuring the MAC addresses in and proceeded to say not to give the same client the same IP address. I also have yet to do the uncomment thing with the drbl.conf file. However, it seems that the DHCP server is going nuts.

                  First, I plugged in a laptop that is not on the mac table and has never logged in before to this server. I booted from network just to test things out and it eventually got in and I was able to image it. How could that happen? I thought that it had to have its MAC recorded unless I uncommented the config file.

                  This seems to really be causing a stir. I continually am keeping an eye on the daemon.log because I am so scared to mess something up on this large network. Every once in awhile I see an unknown computer try to get an IP addy. Here is an example:

                  Jun 26 12:21:58 clonezilla-desktop dhcpd: DHCPOFFER on 10.0.0.36 to 00:19:d2:4e:84:25 (EDEN0457) via eth1

                  I don't know who that is but they tried to get an IP from this server. The log is also continually flooded with messages such as this one:

                  Jun 26 13:07:21 clonezilla-desktop dhcpd: DHCPINFORM from 10.1.15.156 via eth1: unknown subnet for address 10.1.15.156
                  Jun 26 13:07:24 clonezilla-desktop dhcpd: DHCPINFORM from 10.1.15.51 via eth1: unknown subnet for address 10.1.15.51
                  Jun 26 13:07:27 clonezilla-desktop dhcpd: DHCPINFORM from 10.1.15.51 via eth1: unknown subnet for address 10.1.15.51
                  Jun 26 13:12:26 clonezilla-desktop dhcpd: DHCPINFORM from 10.1.15.15 via eth1: unknown subnet for address 10.1.15.15

                  It goes like that for a while. I asked around and this address comes from the other DHCP server. What is going on here?

                  Another problem is I just imaged one of the laptops and then forced the same image onto it to test things out. When I go into the operating system, it continues to be assigned the same IP address from the server. Its MAC is not on the mac table but it continues to go to this DHCP server rather then the normal one.

                  I just need to figure out if there is a way for me to not go in and manually lock down MAC addresses but at the same time let people in when I need to fix a computer and force an image on it. I dunno if I explained what I am doing but we have tons of desktops nad laptops here. I plan on making a perfect install of each one, then imaging them on to the server. Once someone with that model computer has a problem, I would force the "perfect" image onto it through the server. Anyone have any ideas? I am going to try to uncomment the config file now just to see what differences I see.

                   
                  • Steven Shiau

                    Steven Shiau - 2008-06-27

                    You already use 2 NICs ? If so, make sure, say, eth0 is the only NIC connecting to WAN, and eth1 is connected to a network switch which is isolated to WAN and only your client machines connected.
                    The layout is something like this:
                    http://drbl.nchc.org.tw/lecture/20080627_NCHC/drbl-clonezilla.pdf
                    in page 27.

                     
    • Lloyd Carroll

      Lloyd Carroll - 2014-03-12

      Just saw this, don't know if you got it fixed, (for others as well). If you have to boot a second time to get DHCP, remember that nic boot is very quick with respect to powering on the machine. If the switchport you're plugged into is capable of spanning-tree, it will not be active (though lights will be on) for about the time it takes to get into windows. Rebooting will give it enough time. Set your ports for workstations to portfast. ie, for cisco switch, spanning-tree portfast.

       

Log in to post a comment.

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

JavaScript is required for this form.





No, thanks