Ubuntu :: How To Make Grub2 Generate A Grub.cfg File Without UUIDs?
Jul 11, 2010
I want to do everything with disk labels. My /etc/fstab is already set up for labels.How can I tell grub2 to use labels? I need it to stop using UUID for root.
I'm trying to set up an Ubuntu server using 10.04 (64-bit), and running into problems after a couple of reconfigurations. Here's the full story:I initially built the server on a 400+GB RAID5 array, putting everything but swap in one partition. Unfortunately, I needed to repartition, putting / in the first primary partition, swap in the second partition, /var/log in the third primary partition and /home in the remaining space on the fourth primary partition.
However, at this step, I ran into some problems with UUIDs in /etc/fstab and Grub2 (I've used Linux for about 9 years, but I'm new to Ubuntu, and I haven't used UUIDs or Grub2 on Gentoo, yet). Consequently, I made the (probably not smart) decision to move back to the /dev/sdX notation I am familiar with.The problem with that is that now I need even more space on /home, so I've added a Dell Powervault and a Dell PERC5/e SATA card to my server. Now, Grub2 tries to boot from the new RAID array on the Powervault instead of the internal RAID array, so I am trying to move back to the UUID notation so that I don't have worry about /dev/sda being the internal array sometimes and the external array at other times.I don't mind being RTFM'd, but I'm having trouble finding pointers to the documentation explaining Grub2 configuration and the UUID notation. Does anyone have pointers to some readable, concise documentation on configuring this in Ubuntu?
I have /dev/sda with Squeeze and Win 7 on it, and /dev/sdb with Squeeze. I've managed to get Grub 2 to boot from /dev/sdb1, but only by disabling /dev/sda from being a boot option in the BIOS. When it is available to boot, and lower priority than /dev/sdb, grub does not recognize the UUIds of the disks. So, I've disabled it for now and can boot from /dev/sdb no problems. Trouble is I cannot get Win7 to boot. Grub prints:
error: no such device: f0903a3a903a081c error: invalid signature
When I boot into Squeeze and run 'blkid' I can see that:
menuentry "Windows 7 (loader) (on /dev/sda1)" { insmod ntfs set root='(hd0,1)' search --no-floppy --fs-uuid --set f0903a3a903a081c chainloader +1 }
I don't understand how Grub 2 cannot recognize the UUIds. Can Grub 2 to be made to work with volume labels or just plain old /dev/... descriptions? Maybe I should give grub-legacy a go.
I generally have at least 4-5 distro's installed at one time - I always install grub to the boot partition of other distros -that was I control grub from the individual distros - I use ubuntu's grub as the default boot loader.
Rather than the method that ubuntu uses with the update-grub.I normally edit the grub.cnf file to add other distros but when update-grub launches it also adds links in grub.How can I prevent the upadte-grub script automatically adding lines for other distros i.e - I just want it to add ubuntu entries and ignore other partitions.
how do I tell grub to stop trying to boot based on UUIDs? I've been dealing with this for about two years now. I have this one system, a PowerEdge 1550, that for some reason WILL NOT boot using a UUID, and will only boot if I edit the first entry in grub, remove everything to do with UUIDs (completely delete the search line, replace root=uuid=[uuid] with root=/dev/sda1). I can't seem to make these settings stick and now with 10.10 there's no longer a menu.lst to edit. So, in short, I need grub to stop trying to boot based on a UUID. Absolute paths, baby. That's what I need.
today I updated my Slackware 13.0 to 13.1. Unfortunately it didn't boot any longer. I probably forgot to run "lilo" before restarting. However, I reinstalled Slackware and I installed grub instead of lilo I'm more familiar with it..
This was my first menu.lst config:
Quote:
title Linux (uuid) root(hd0,0) kernel /vmlinuz root=UUID=6a8096a3-3915-4ef2-8984-976e42d04cfc ro vga=0x031b
While booting it stopped and printed this message:
Quote:
Kernel panic - not syncing: VFS; Unable to mount root fs on unknown-block
Then I compared it to other Linux system on different computers and I noticed that every of them, who are using uuids, using an initrd as well. So I created one and now it boots properly.
I just wonder, why? Why does it need an initrd to boot by using uuids?
make working entries in grub.cfg. This is taken from Ubuntu 10.10 and despite the do not edit warning, I edited it with sudo gedit /boot/grub/cfg. My computer boots into every OS in the list.
# DO NOT EDIT THIS FILE # # It is automatically generated by grub-mkconfig using templates # from /etc/grub.d and settings from /etc/default/grub #
I have (among others) these partitions: /dev/sda5 'newsys' - my current system partition (Ubuntu Karmic, 64-bit) /dev/sda2 'oldsys' - the original system partition /dev/sda1 'boot' - mounted at /boot when running 'newsys'
I have ubuntu 10.10 running on my HP dv5 pavilion laptop So today i tried to reinstall grub on my pc .. i removed grub-pc and grub-common using synaptic then booted up with a live cd of ubuntu 10.04 i then mounted sda11 to mnt using sudo mount /dev/sda11 /mnt sda11 being my ubuntu directory containing the /boot then i installed grub to it using sudo grub-install --root-directory=/mnt /dev/sda it reported as installation finished.No error reported then i unmounted /mnt and checked for /boot/grub/grub.cfg seems like the file is missing so are the files in /etc/grub.d/ reinstall my grub back
I have a "master" IDE HD upon which reside several Linux OSs and a couple of Windoze OSs. The boot system is Mandriva using Grub Legacy and the latter's menu.lst file is where I select which OS I wish to use. I use HD physical carriers for my HDs. I used GPartEd to copy over newly installed OSs on another drive to the "master" HD. The commands in the menu.lst file for Linux OSs are representative as follows:
Name of OS root (hdx,x) configfile /boot/grub/menu.lst
Task: Enter Grub2 which, I seem to understand, is a work in progress. I presume that some commands in Grub2 are not recognizable by Grub Legacy. I installed Debian 6 on another drive and I wish to incorporate an entry in the menu.lst file described above using the technique described above. Is it possible to use the menu.lst file in Grub Legacy (0.97) in order to boot the Grub2 Debian OS? Or am I chasing windmills?
I have had it with the frustration of trying to figure out how to successfully install grub on my own, so I am hitting the forums. What can you tell me about installing grub/grub2 (whichever is best -- I want something that works) to my single hard drive computer. Operating systems: Windows XP, xUbuntu 9.10 Hard drive: Samsung SATA 2, 1 TB Desired result: Menu that loads upon boot so that I can select which OS I would like to load without inserting media device or disk.
I thought this individual had the most straight forward and thorough tutorial (of any I have seen) on the subject but precluded at least one thing for me that I cannot afford to ignore: I don't seem to have a "stage1" that can be loaded by the grub shell. Trying to find it in /boot/grub/stage1 in the grub shell, I get... "Error 15: File not found" What I do have in that directory is 'grubenv'. When I boot up, it loads grub, but it comes up with an error message saying, "File not found". Then, all I get is the 'grub rescue' prompt: "grub rescue>".
Obviously, after spending hours on this thing, this is not the whole story. But there have been so many attempts to try fix it with one thing or another, that I cannot easily relate them all, if I could even remember each one.
I was tried to compile the brl-cad for my IA32 machine. I install all the GNU auto series (autoconf, automake, libtool). Then I run sh autogen.sh asnstructed by the installation guide. But after the preparation finish, I check my directory. The make script does not exist. And everything broke down here. When I tried to run "make", I got this feed back: make: *** No targets specified and no makefile found. Stop.(of course, there is no make file)I had run every step by using root account. Hence, there should not be any restriction on operation.
Previously I have Centos 5.3 installed. Eventually I upgraded the systems to Centos 5.4 Now is it possible to generate an ISO and burn into a DVD to make a new copy of Centos 5.4 installation disk? This is far better than downloading.
I have an old BIOS (only 149 GB detected in 160 GB drive). I have installed very few OS as a result - Windows in first 80 GB , 2 distros (20 GB each) in next 40 MB which is within the 149 GB. On issue of "update-grub" grub seems to detect all OS and generates grub.cfg. But on boot , one distro does not appear in the menu.
So I use the following to rip audio from one .mp4 file and then generate one .mp3 file.
Code: caleb@linux-vmp1:~> ffmpeg -i a -ab 128000 -ar 44100 01.mp3
Obviously, this takes a file named a and makes a file named 01 that is an mp3. It works great and I have used it many times. The problem is, I would like to rip the audio from several files all at once. I am sure there is a way to do it but I haven't stumbled upon the way to do it.
I'd like to use a keymap not present in slackware "fr-dvorak-bepo" with mkinitrd because my root partition is encrypted by luks. So how can I generate a .bmap file from a classic .map file for mkinitrd ?
After upgrading to 10.04 from 9.10 Win7 wouldn't startup any more. So I tried this HowTo: [URL] to restore Grub2. But now each time I boot up I get this two lines: error file not found grub rescue> I have NO idea what to do.
Since the computer does not have xorg.conf, can I have the computer generate that file because I need to work on it and so I need to create it. That is why it would be useful to have a xorg.conf file still.
I'm trying to write a script to generate an html file (complete with formatting- echo "[random formatting]" >> index.html) for all the files in the given directory. So far, it works pretty well. HOWEVER, I want the listed files to be treated as links. I'm using awk to grab the part of the filename I want, but I don't know how to do this out as it fails if there is more than one file. The HTML side would look something like this:
Code: <li><a href="2010.05.29.html">May 29</a></li> It all works fine up to the actual number of the day- fine with one file, fails with more than one. My code is this: Code: # Grabs all the files and puts them in a list with anchor text "Listed" ls | find 2*.html | sed -e 's/^/<li><a href="/'
I'm trying to learn how to create a user account manually on the system, and I've edited the /etc/passwd and /etc/groups as well as creating a new home directory by copying /etc/skel but I'm stuck at how to generate an entry in the /etc/shadow file since it comprises of the hash and all?
I'm having trouble with this developing tool... I dunno why but when I build my program, Lazarus compiles it, but keeps debugging and don't generate the file. Do anyone knows which may be the problem?
I have an old laptop that I installed EasyPeasy Lucid on. I have never used Grub2 as my other Ubuntu installs still use Grub Legacy. I have studied all the tutorials which instruct that the grub.cfg file should not be edited directly.This old laptop has one of the dreaded Intel graphics chips for which the kernel automatically loads the i915 module. Of course (like a multitude of others with Intel graphics and Lucid) I booted into a black screen but knew the workaround was to enable mode setting through grubI used the
Code: i915.modeset=1 and added it to /etc/default/grub line GRUB_CMDLINE_LINUX, just after "quite splash", just
I'm using RHEL 5.4 and trying to use the system-config-kickstart to generate a ks.cfg file with all the settings already appeneded. After running the "system-config-kickstart --generate ks.cfg" command, the file gets created but it's missing the firewall configuration, partition information and so on.
How can these settings also be generated with the system-config-kickstart?
I am using RedHat MRG on a standalone intel PC that has 8 cpus. It has 2 Quad cores. I am trying to force some specific persistent tuning on it so that the OS and IRQs are all on one CPU socket and then I can run my Application on the other CPU socket (4 CPUs) using the taskset command.
I have been reading up on this mainly at [url] and having some success. But, I have one IRQ that is not active until my application runs that I want to force it to be on CPU 4.
And so I believe I can do this by setting /proc/irq/default_smp_affinity to 10. But, I do not have such a file on my system. How can I create this file at boot time properly?
I installed Ubuntu 9.10 yesterday on my computer. I had Windows XP and Windows 7 (I deleted the Windows 7 partition and installed Ubuntu 9.10). However, after installing Ubuntu, I couldn't access the grub boot menu. So I googled, and I found out how to make the menu show up again. That was okay. But Windows XP didn't show up. So I googled a bit more, searched the forums, and I added the following entry to my 40_custom file code...
As far as I know, that should work. The uuid I entered there is the one that gparted told me was the uuid of the partition. However, this isn't working. I get a disk read error when I try to boot my Windows XP.
According to my gparted, my Windows XP drive is inside an extended partition? I wonder if that has any influence on why I'm getting the disk read error, or anything. It's either that, or my 40_custom entry is wrong. I've been trying to sort this out for a while now, and I'm afraid I'm not getting anywhere with it.