My evil nemesis Grub2 has decided to make a move on me again! So I rebooted a few minutes ago and got this:
Welcome to Grub
error the symbol 'grub_env_find' not found
Entering rescue mode
how to use Grub2 to boot Windows XP and Linux.
[code]...
there were no entries in the Grub 2 menu after booting. I don't really understand the numbering of hd0,0 to like sda, but I changed "root (hd1,0)" to "root (hd0,0)" if Windows is on hd0, but after rebooting, I still didn't see it. I read somewhere that Grub2 starts partitions at 1, but wasn't sure that was true because in menu.lst they start with 0 sometimes. know entry-level steps to solve this?
I upgraded a while back to Windows 7. I run two separate HDDs, with debian on sda and windows on sdb. Everything worked fine, up until i upgraded to windows 7. The installer forced me to unplug sda. After installing I repositioned the HDDs and was stuck with a "NTLDR is Missing" message. After removing the search -fs--uuid tag from grub.cfg, Windows 7 booted for a good 0.5 sec and then rebooted. All that flashed was the "Starting Windows" page. I see similar bugs reported in the launchpad, but nothing that directly parallels this.
Here's my windows entry in grub.cfg:
Changing ${root} to (hd1) makes no difference. Also, I've been using Windows 7 for moths now without a problem, it's only when I try and boot it form grub2 that I get a problem. Physically removing sda lets Windows 7 boot fine.
I recently upgrade my grub to grub2 and had problem to boot debian. After trying to fix it myself, I played around with grub-image-ntldr, my debian now works ok, but I cant boot my XP anymore.
Now when I try to boot XP, I got "No grldr" message. If I copy my ntldr and rename it to grldr, my XP will boot ok. So it seems that when I try to boot XP it doesn't chainload ntldr anymore, it is looking for grldr.
I tried to change "chainload +1" to "chainload /ntldr", it complains about invalid signature. Does anyone knows why my grub now using grldr instead of ntldr?
I Upgraded the system to 2.6.26-2-i686 doesn't boot:
error message on Grub is:
I have a ~ 2008 notebook (Compaq CQ60-137EL) on which I had Windows 7 only (it was sold with Windows Vista installed).
Later I installed Debian Jessie 8.2.0 Stable ("Graphical expert install" from DVD), along with GRUB as a boot manager (I chose not to install it on the EFI removable media path).
Since then, if I select Windows 7 on the GRUB boot screen, I see "Starting Windows...", and after few seconds the screen flashes for a moment, and then the PC reboots: I see the bios screen, followed by the GRUB screen again.
What's even more weird about this is the fact it just happens only in like ~50% of the cases. In the other 50%, Win7 starts flawlessy.
I even tried to install Debian first, then Windows 7, then re-install GRUB, but I got the same issue, even with both system freshly installed.
On 6 attempts, 3 times it worked and 3 times it didn't.
On my desktop PC I'm in the same setting, but I don't have this issue. I think it may be related with the fact I have Win7 on a SSD and I installed Debian on a separate HDD, while on my notebook, as you can imagine, there's just one single HDD.
I have just recently installed firestarter to get an idea of iptables editing with a GUI. Firestarter works great and I love it, however when I turn my system on, and my system is booting up the terminal displays a "Failed" message when attempting to start up firestarter. Now I have been doing my homework (reserch) and have read that a user will get this error because firestarter trys to load prior to the network manager starting up. I have also read some other variables that can cause this is if you use network manager and have a password on your user account (which I do have) can cause this issue.
However in most cases from what I have read is even though it gives the user a "Failed" startup message, once the user logs in, and is connected to the internet firestarter will actually load the iptables. I did some tests and from what I understand I can see that this is true because I have allowed access to certain torrent trackers, and denied access to others, and I can see some come online. Does firestarter load iptables once the user logs in? Or once it fails to load, does it not load at all unless you manually open/load the application. If firestarter just modifies iptables, then once I save my listings in firestarter then I really don't need it to start up correct?
I recently installed Ubuntu Netbook Remix (10.04) on my brother's Acer Aspire One (751h) netbook, set up to dual boot with XP Professional. The initial install was successful, with both OS's booting correctly. However, after a series of updates and module builds (this machine has the Poulsbo chipset, still somewhat of a mess), Grub2 is now unable to boot XP, displaying the error "Operating system not found."
To be clear, grub works properly (aside from not booting XP), so the MBR is intact, although I've rerun grub-install to be safe. Restoring the windows MBR (using testdisk) boots XP happily, but of course ignores the ubuntu installation. I also had testdisk rebuild the NTFS boot sector, to no change - grub still doesn't work, the windows MBR does. Using the grub console, I can "root (hd0,1)" and see files on the XP partition, but "chainloader +1; boot" again fails with the same error. The ubuntu install can likewise mount the NTFS partition with no problems.
Given the fact that switching the MBR fixes the issue, I'm suspecting a Grub2 configuration problem. However, the simplicity of "root (hd0,1); chainloader +1; boot" leaves me stumped as to what that could be. Also, the error text "Operating system not found" is suspiciously identical to my BIOS's error message when attempting to boot from a non-bootable medium (as I found out while attempting to make a bootable usb stick with my mac).
instead of normal grub menu I am greeted with an:
Code:
symbol "grub_env_export" not found
on debian_amd64 latest testing.
[code]....
I have a Windows Vista-Ubuntu 11.04 dual-boot. However, I have lost the ability to boot Windows Vista. I can select Windows in the GRUB menu, but GRUB says
Quote:
error: symbol not found: 'grub_err_printed_errors'
I have repeatedly searched Google for any help, but have found nothing that applies to my situation. I need Windows because I have an iPod Touch with a tethered jailbreak and can't use it right now.
After removing a hard drive I (thought I)wasn't using, GRUB failed to load(turns out stage1 was on that drive) and it refused to install to a new drive(even after I kexec'd into the system - which was fun, considering the LiveCD used a different name for the hard drive). I finally threw in the towel and installed GRUB2, which worked after removing a second, incorrect root=. However, I can't find out how to switch the font from the fugly default to something that doesn't try to gauge my eyes out with a rusty spoon.
View 1 Replies View RelatedI have just upgraded from Lenny to Squeeze, and followed the D-I recommendation to upgrade to Grub2. However, the server will no longer boot - giving a 'missing operating system' error, so it's not getting very far in the boot loader process. I have read, in various location that Grub2 supports GPT, but my efforts to fix the situation have met with little success.
I have managed to boot the system using SuperGrub2 (recovery live cd) and can start up successfully from there. However, when I try to fix the grub installation using the Debian wiki recommendation, I'm met with:
<host>:/# grub-install /dev/sda
/usr/sbin/grub-setup: warn: This GPT partition label has no BIOS Boot Partition; embedding won't be possible!.
[code]....
The attached disk is 4.5TB of RAID 5 disk behind an Adaptec 5405 if that's useful.
Daily updated Debian Testing Because Debian is the only operating system on this laptop and I keep at least two working kernels, I would like to hide Grub2's menu unless I press a key (like one could do with Grub). I can hide the menu if the line GRUB_TIMEOUT=0 is in /etc/default/grub but it doesn't appear after pressing SHIFT, which is a threat if the system cannot boot the selected kernel. Right now, the timeout is set to one second. I've read Grub2's documentation and [URL] and tried various combinations, but I haven't been able to make this work.
View 1 Replies View RelatedI just got a new hard drive and figured I might as well do an installation of Squeeze (and was previously using Lenny). That went fine, and then I decided that I should copy over my old Lenny installation to the new disk -- mostly to have a working backup without bothering to do a new installation on a partition of the new drive. My partition scheme was a smaller /boot partition and then a much larger / partition with everything else standard (and a much larger /data partition rather than storing everything under /home). So I copied /boot over to a new partition on the new disk, and the same with /. That was done from the Squeeze installation, so the Lenny install wasn't active at the time. I modified all the appropriate entries in /etc/fstab to use UUIDs rather than partition numbers and ran an update-grub.
It detected everything on the old and new disks without a problem. When I went to try and boot up the transferred Lenny installation, it hung on trying to activate the root file system (I've forgotten the exact messaging). Not entirely unexpected, mind you. I went and took a look at the grub.cfg file. It does list that the transferred Lenny is on partition sda8 (correct), it has the correct uuid for the boot partition... but it seems to be setting the root incorrectly. Specifically, the root is still set for the old disk (though in its new position of hd1 instead of hd0), and the "linux" line sets a root for the old device. Or more specifically, this is the menu entry I get, with a few // comments.
menuentry "Debian GNU/Linux, kernel 2.6.26-2-amd64 (on /dev/sda8)" {
insmod part_msdos
insmod ext2
set root='(hd1,msdos2)' // <------ This position is the *current* location of my old Lenny disk/partition
search --no-floppy --fs-uuid --set [the correct /boot UUID]
linux /vmlinuz-2.6.26-2-amd64 root=/dev/sda5 ro vga=795 // <----- that root=/dev/sda5 line is what it was on the old device.
initrd /initrd.img-2.6.26-2-amd64 }
The odd thing there is that it sets the root according to the *current* position of the old install disk (maybe some trickery with detecting the correct UUID before setting the root), but the "linux" line refers to the *old* partition. The two lines will never match up no matter what. Now if I edit grub.cfg manually, I can make it boot (and run) normally, as I've verified. For example, I made a couple manual changes to do this:
menuentry "Debian GNU/Linux, kernel 2.6.26-2-amd64 (on /dev/sda8)" {
insmod part_msdos
insmod ext2
set root='(hd0,msdos3)' // <----------- First hard drive, partition sda3 is where I put /boot
search --no-floppy --fs-uuid --set [the correct /boot UUID]
linux /vmlinuz-2.6.26-2-amd64 root=UUID=[the correct / UUID] ro vga=795 // <--- Changed it to the UUID here; could be /dev/sda8 instead
initrd /initrd.img-2.6.26-2-amd64 }
And that works perfectly, the installation works just fine. Of course, the problem is that those changes will/would be lost every time I run an update-grub. So my question, in a nutshell, is how can I configure things so that update-grub sets things properly to the new devices? Or in other words, where in the copied installation are the variables I need to change? I did note one thing odd -- the existence of a vga=795 line. For the new Squeeze installation, I'm using gfxpayload and there's no vga=anything line anywhere. My old installation, of course, had its own grub installation where I did use vga=795 to set the console resolution properly.
So my first guess was that update-grub (for Grub2) was pulling config information out of the Lenny /boot/grub folder (grub legacy). Unfortunately I tried several changes there and it made no difference. Then I deleted the entire /boot/grub folder entirely from the copied Lenny installation and ran update-grub again (the Squeeze grub version). It changed absolutely nothing. That's very confusing for me, since I have no clue where it could be getting vga=795 from, *except* the now-deleted Lenny /boot/grub folder. Where in the copied Lenny installation, I can change something to make it so that update-grub picks up the correct information?
After an update on squeeze about a week or two ago, my console font turned green. I'm not exactly sure of the timing, because I switch between lenny and squeeze, and some time passed between the update and the reboot in squeeze.
During the update, I was asked if I wanted to keep my altered /etc/grub.d/00_header or take the package maintainer's version. I took the PM's version, knowing I'd lose my edit (set gfxpayload=1024x768x16). When I add that line to the new 00_header, run update-grub and reboot, the console font is the size I want, but it's a dull green. Reminds me of my first computer. How do I get it back to white and keep the 1024x768 resolution?
Currently booting with grub-legacy and chainloading grub2. If I boot entirely with grub-legacy, and pass vga=791, the console font will be white. It doesn't matter which kernel I use. Currently, I'm running 2.6.32-3-686. Same behavior exists if I use 2.6.30-2-686. If I use 1024x768x8 I get a dull gray instead of dull green.
I'm using Truecrypt to encrypt my Windows 7 OS. I also have unencrypted Ubuntu 10.04 installed on /dev/sda6 on the same hard drive. Since Truecrypt bootloader must be installed in MBR, I have GRUB2 installed on /dev/sda6, so I can use TC bootloader to load GRUB2. When I first install GRUB2 on /dev/sda6, I can use TC bootloader to load Ubuntu. But, if I boot into Windows via TC bootloader, and then later try to boot into Ubuntu, I get the message "no bootable partition found". I have to reinstall GRUB2 onto /dev/sda6, every time after I use windows in order to be able to boot into Ubuntu. It seems that starting Windows somehow overwrites GRUB2. Is there a fix for this?
View 1 Replies View RelatedWe have a SuSE/SLES 9 server that boots from a fibre channel card, using volumes from a NetApp filer.
We previously had the server booting from the LUNs on the filer, but after some tinkering around with the fibre channel BIOS, we have the situation where the LUNS seem to be mounted, the OS boots, starts to initialise everything but then stalls, saying:
We have tried loads of combinations of settings in the BIOS and fibre channel BIOS without any success or any idea what may have caused the error.
I've been installing multiple versions of ubunutu over years and now I better understand why I faced always problems with wifi configuration on these different baselines : Debian root of course.
Now on Debian 7.4 :
The Broadcom BCM4313 driver is not loaded by default (not free product) for my wifi card. This is stated thru the UNCLAIMED declaration
Code: Select allÂ
sudo lshw -c network
 *-network UNCLAIMED
    description: Network controller
    product: BCM4313 802.11b/g/n Wireless LAN Controller
    vendor: Broadcom Corporation
    physical id: 0
[Code] ...
Unclaimed means : no driver found (no kernel association)
So I installed the driver located in the non-free backports by adding this target into the repository
Code: Select alldeb http://http.debian.net/debian/ wheezy main contrib non-free
and install :
Code: Select allsudo apt-get update && apt-get install firmware-brcm80211
sudo modprobe brcm80211
Check seems OK :
Code: Select allsudo lshw -c network
 *-network
    description: Wireless interface
    product: BCM4313 802.11b/g/n Wireless LAN Controller
    vendor: Broadcom Corporation
    physical id: 0
[Code] ....
I can't connect with wlan0, it doesn't work. I found information here regarding WPA2/PSK wifi configuration on debian : [URL] ....
so I added this setup to my interfaces file an got :
Code: Select allauto lo
iface lo inet loopback
allow-hotplug wlan0
iface wlan0 inet dhcp
 wpa-ssid Hznteam-Datacenter
 # hexadecimal psk is encoded from a plaintext passphrase
 wpa-psk 12345678901234567890123456
It's worth ! no more wifi network are detected and ifconfig gives no ip address allocated :
Code: Select allwlan0   Link encap:Ethernet HWaddr 78:e4:00:4e:49:a3Â
     UP BROADCAST MULTICAST MTU:1500 Metric:1
     RX packets:0 errors:0 dropped:0 overruns:0 frame:0
     TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
     collisions:0 txqueuelen:1000
     RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)
[Code] ....
I've just installed lenny from dvd, and am just settling in. I'm curious about virtual machines. so I've installed xen. With xen comes qemu by default. Now the setup I've recently become familiar with is Suse-11 which is quite slick.In Suse, Yast provides a distribution prepared virtual management section for xen, including an installer. Debian's setup is similar but not a clone. I want to try out the installer. Aside from the differences, debian has a screaming deficit. On booting dom0 in Lenny, if fails to get into gnome, badly. I can tell what's probably wrong, but I need to fix it. Now I'm also a gentoo fan and very practised at posting queries, so here goes.
Debian's config for X is somewhat unhelpful. The xorg.conf just states Configured device and configured monitor for its components. If it expanded on just what the settings were it would help. What I have is a frambuffer driver not loadinf or being implemented.
I have squeeze using the default setup with 5 seperate partitions on hard drive. All other updates are fine but it will not install the Linux-Image-2.6.32-5-amd64. I get message that not enough room. I think this is the kernal update. I use update manager. Is there a fix for this short of resizing the partition. This is a one computer/user setup. I had this problem on a previous install of stable and ignored it thinking that it was a glitch. That install was an experimental setup and I liked it but it didn't work on the internet very well so went to testing.
View 5 Replies View RelatedI have happily been booting debian through grub2 by chain loading it with efi (rEFIt), until today, and now get to begin another learning experience I've been using linux for a while, and kept seeing the guides for splitting up /, /var, /tmp, /usr, and /home, into different partitions, so I did just that when I switched from Ubuntu to Debian (I've realized that this was a little bit pointless because I formated them all as ext4, but at least it acts as a safety for mission critical drives when I overfill /home. I unfortunately didn't give /tmp enough space, and it kept crashing SimpleScan so I decided to use gparted to resize it.
The operation went alright as far as I can tell, and was straight forward because there was some free space behind it so I only had to append the partition. I synced the master boot record through rEFIt as usual, but when I booted the linux partition grub did load, and only a blank screen is presented. I eventually figured out I could use the gparted live cd to boot back into debian, and have been screwing around for a while with grub commands trying to figure out how to allow rEFIt to successfully boot GRUB on its own again. I ran grug-mkconfig to replace my /boot/grub/grub.cfg file and have rebooted but that did not help.
I tried reinstalling grub and grub-common with apt-get, but I didn't purge configuration settings for fear of losing something important. My current focus is on the command grub-install. I think i just need to run this command with the /boot device, like su - root; grub-install /dev/sda1 or some thing like that. wipe out the MBR on /dev/sda1, or screw up what good configuration is left in grub, so I want to make sure that I'm using the right /dev. Currently the gparted output looks like this:
/dev/sda1: fat32 - GPT (gpt from fdisk, gparted shows EFI with the boot flag)
/dev/sda2: hfs+ - MacOSx
/dev/sda3: ext4 - /root
[code]...
how the gnome live gparted disk would have been able to boot. I have access to a hard drive so I'll probably end up making backup images of as many of the partitions as I can, and then try more drastic bashing around, but if anyone has any suggestions/wisdom they could offer while I'm researching solutions I'd appreciate it. I eventually want to try to axe my osx partition and boot directly from GRUB2-EFI so I figure it is worth the investment in time to get to know grub a little bit more intimately.
My debian machine works fine other that it loses network after a few (12 or so) hours. The dhclient manages to renew the DHCP lease the first time, after a few hours, but later when it tries again the modem gives a DHCPNAK. If I understood correctly, this should make the dhclient to fetch a new ip address but it does nothing. I am no expert so I don't know if the cause is the modem box or the debian. There are other (windows) machines connected to the modem box which work fine. The machines are behind NAT.
The debian machine does not currently have a display or a graphics card so it's difficult to start messing with the settings. I've considered having a static ip but I don't understand why the dynamic address doesn't work. What is weird is that it worked fine for several weeks and suddenly started failing. I was running a minecraft server (hehe) with an uptime of several weeks. At first I though it was a hardware issue because I thought the whole computer froze but instead it just drops the connection and I can no longer reach it with SSH or anything.My timezone is UTC+3, /etc/default/rcS has UTC=yes, I believe it is the reason why the lease and the logs differ by three hours.The machine is only running apache2+mysql+php5, GDM is stopped.I'm trying to post as much info I can.Here is the ifconfig output when the connection is working for the wired connection:
eth0 Link encap:Ethernet HWaddr 00:25:22:2b:28:f6
inet addr:10.0.0.7 Bcast:10.0.0.255 Mask:255.255.255.0
inet6 addr: fe80::225:22ff:fe2b:28f6/64 Scope:Link
[code]....
I am using Jessie with XFCE and I tried to write a service which executes a script to change my wallpaper. When I try to start the service with systemctl start wallpaper.service it fails and I get the outout below from systemctl status wallpaper.service
I don't think it to be a permissions issue, they are -rw- r-- r--
This service is called by a timer that goes off daily. Below is wallpaper.service
I finally have my system to where I can connect to proftpd but my uploads fail. I think my problem lies in the upload folder. I do not really understand where to put it. Will someone look at this config file and tell me where to put it and what I need to do to get it working. A basic anonymous configuration, no upload directories.
[code]...
I would like to use debian's popularity contest utility as I think it's an important tool for the developers and packagers, but it seems to be broken. I received the following notification after a recent full-upgrade from Lenny to Squeeze:
From MAILER-DAEMON Thu Nov 18 19:14:25 2010
Envelope-to: root@mediacenter.local
Delivery-date: Thu, 18 Nov 2010 19:14:25 +0200
X-Failed-Recipients: survey@popcon.debian.org
Auto-Submitted: auto-replied
[Code].....
I'm having problems with resume after suspend to RAM. The machine starts to wake up, but the screens (multi mon VGA and DVI setup) are black and the keyboard doesn't light up. After ~20 seconds there's some brief disk activity and then the computer reboots. 100% repeatable with affected kernel versions. My test method is simple, I boot the machine on the kernel's recovery option, log on as root and run "PM_DEBUG=1 pm-suspend". I haven't found anything in the logs after a failed resume.
Here's the situation:
I have a SSD disk. To get TRIM support I have to use kernel 2.6.33 or later, which means that the standard kernel in Squeeze is too old.I have Nvidia graphics, and there was a change in 2.6.34 that breaks older versions of xserver-xorg-video-nouveau (version 0.0.15, used in Squeeze), I can't use Debian Squeeze with a kernel newer than 2.6.33.x.My machine (XFX GeForce 9300 motherboard) won't resume from suspend to RAM if I use a kernel newer than 2.6.36. There are no BIOS updates available.
My options:
Install newer kernel from Squeeze backports (2.6.38.2 last time I tried). <--- Not doable b/c of resume problems. Upgrade to Wheezy, which uses kernel 3.0.x. <--- Not doable b/c of resume problems.Compile a vanilla kernel. So basically I'm forced to compile my own vanilla kernel, 2.6.33.x on Squeeze or 2.6.35.x on Wheezy. I won't be stuck with an unsupported kernel version in the near future, but so far I've failed miserably.
I know that the latest kernel version where everything works is 2.6.36.x (no longer maintained), 2.6.37.0 and later cause resume problems (I've tried 2.6.37, 38, 39 and 3.0.0, .0.1). I've tried doing a git bisect on the kernel, but didn't succeed, ended up on 2.6.36-rc5 which is weird considering that 2.6.36.4 works. There may be several suspend/resume bugs in different kernel versions that messed up the bisecting results.
I am running Jessie 8.3 and wireless in Wicd has stopped working for me, displaying "No Wireless Networks Found."
Code: Select allsudo /etc/init.d/wicd restart Restarted wicd
Made sure wlan0 is in wicd preferences.
iwconfig:
Code: Select alleth0Â Â Â no wireless extensions.
wlan0   IEEE 802.11abg ESSID:"Tell My Wifi Love Her"Â
     Mode:Managed Access Point: Not-Associated  Tx-Power=200 dBm Â
     Retry short limit:7  RTS thr:off  Fragment thr:off
     Encryption key:off
     Power Management:off
    Â
lo    no wireless extensions.
For some reason my user account was removed from the netdev group? So I readded it...
Code: Select allekarr@fidelio:~$ sudo gpasswd -a ekarr netdev
Adding user ekarr to group netdev
[code]....
And still, "no wireless networks found" in Wicd.
my system keeps on telling me about an error:
Code: Select allsudo journalctl -p 0..3
Apr 14 10:36:11 debian ntpd_intres[682]: host name not found: ptbtime1.ptb.de
Apr 14 10:36:11 debian ntpd_intres[682]: host name not found: ptbtime2.ptb.de
Apr 14 10:36:11 debian ntpd_intres[682]: host name not found: ptbtime3.ptb.de
A quick check
Code: Select allsystemctl status ntp.service
ntp.service - LSB: Start NTP daemon
  Loaded: loaded (/etc/init.d/ntp)
 Â
[code]...
3 packets transmitted, 0 received, 100% packet loss, time 1999ms.Same result when using the standard Debian ntp time servers like "server 0.debian.pool.ntp.org iburst" in /etc/ntp.conf.
Installed linux-image-2.6.32-bpo on my laptop today and after the reboot i got "no operating system found" . I'm now typing using a Puppy live cddrom .
View 3 Replies View RelatedI'm setting up the Squirrelmail | impad | sendmail combination and so far everything appears in order but I'm getting a 'mailbox not found' error when I log into Squirrelmail. I've eliminated Squirrelmail by using another imap client (Mozilla Thunderbird) and getting the same error. It seems that the problem is with imapd.I've checked the logs and I can see only the successful login entries. Nothing else:
May 5 10:30:19 altair cyrus/imap[2602]: executed
May 5 10:30:19 altair cyrus/imap[2602]: accepted connection
May 5 10:30:19 altair cyrus/imap[2602]: login: localhost [127.0.0.1] john plaintext User logged in
[code]....