"Unable to migrate to dependency-based boot system..." - Forums Linux |
- "Unable to migrate to dependency-based boot system..."
- Weird ubuntu audio/sound problem
- How do I make resize my /'s /boot partition for more free disk space?
- Where is Bluetooth installation coming from and how can I removeit from my old Debian box?
- cybercafe program
- Getting *all* local mail to be sent to the smarthost relay
- Issues with 7TB partition
- Can my ubuntu desktop access web via wirelessrouter
- Newbie looking for advice on Linux distribution
- Protecting the Windows using Linux
- Intel Graphics Under Debian Lenny (Blank Screen)
- cannot change current block/inode allocation
- GUI for /etc directory?
- Pimp my GRUB
- install linux on blade server
- SQUID problem;need help
"Unable to migrate to dependency-based boot system..." Posted: 01 Mar 2010 04:33 AM PST On 3/2/2010 5:40 AM PT, J G Miller typed: Thanks man! I will wait for the Squeeze. I will do like I did last time and use net-install. Amusing how my XP Pro. installation is older than this Debian box. I need to redo Windows too with 7 one of these days. :) -- "Ants die in sugar." --Malawi /\___/\ / /\ /\ \ Phil./Ant @ http://antfarm.ma.cx (Personal Web Site) | |o o| | Ant's Quality Foraged Links: http://aqfl.net \ _ / Nuke ANT from e-mail address: netANT ( ) or com Ant is currently not listening to any songs on his home computer. |
Weird ubuntu audio/sound problem Posted: 28 Feb 2010 10:29 AM PST edu wrote: Did you make the mistake of installing the OPTIONAL modem driver? When you point to your Volume Icon, you should get a Tooltip with something like "Internal Audio og Stereo IF YOU DID install the modem, uninstall it and see what happens. That fixed this problem when I upgraded to Ubuntu 9.10 That's because my on-board Modem is a Voice Modem used for Voice-over-IP used for using my system's speakers & mic for phone function. You don't want the feed-back squeal when you use it as a phone, hence the speaker is disabled with the modem installed. |
How do I make resize my /'s /boot partition for more free disk space? Posted: 25 Feb 2010 02:50 PM PST Woohoo/W00t! I managed to install Kernel v2.6.32 barely on my old Debian box: # apt-get install linux-image-2.6.32-trunk-686 Reading package lists... Done Building dependency tree Reading state information... Done Suggested packages: linux-doc-2.6.32 The following NEW packages will be installed: linux-image-2.6.32-trunk-686 0 upgraded, 1 newly installed, 0 to remove and 127 not upgraded. Need to get 26.3MB of archives. After this operation, 76.0MB of additional disk space will be used. Get:1 http://ftp.debian.org testing/main linux-image-2.6.32-trunk-686 2.6.32-5 [26.3MB] Fetched 26.1MB in 35s (731kB/s) Reading package fields... Done Reading package status... Done Retrieving bug reports... Done Parsing Found/Fixed information... Done Preconfiguring packages ... Selecting previously deselected package linux-image-2.6.32-trunk-686. (Reading database ... 158155 files and directories currently installed.) Unpacking linux-image-2.6.32-trunk-686 (from .../linux-image-2.6.32-trunk-686_2.6.32-5_i386.deb) ... Setting up linux-image-2.6.32-trunk-686 (2.6.32-5) ... Running depmod. Running update-initramfs. update-initramfs: Generating /boot/initrd.img-2.6.32-trunk-686 Running update-grub. Generating grub.cfg ... Found background image: moreblue-orbit-grub.png Found linux image: /boot/vmlinuz-2.6.32-trunk-686 Found initrd image: /boot/initrd.img-2.6.32-trunk-686 Found linux image: /boot/vmlinuz-2.6.30-2-686 Found initrd image: /boot/initrd.img-2.6.30-2-686 done Examining /etc/kernel/postinst.d. run-parts: executing /etc/kernel/postinst.d/initramfs-tools 2.6.32-trunk-686 /boot/vmlinuz-2.6.3 2-trunk-686 localepurge: Disk space freed in /usr/share/locale: 0 KiB localepurge: Disk space freed in /usr/share/man: 0 KiB localepurge: Disk space freed in /usr/share/gnome/help: 0 KiB localepurge: Disk space freed in /usr/share/omf: 0 KiB localepurge: Disk space freed in /usr/share/doc/kde/HTML: 0 KiB Total disk space freed by localepurge: 0 KiB $ df Filesystem 1K-blocks Used Available Use% Mounted on /dev/hda1 280003 240465 25082 91% / .... I probably won't be able to install 2.6.33 even if I were to uninstall 2.6.30. :P When that happens or whjen my hardware fails or whatever, then I will do a clean install from scratch and do my partitions correctly. Thank you to all who replied. :) -- "We are anthill men upon an anthill world." --Ray Bradbury /\___/\ / /\ /\ \ Phillip (Ant) @ http://antfarm.ma.cx (Personal Web Site) | |o o| | Ant's Quality Foraged Links (AQFL): http://aqfl.net \ _ / Please remove ANT if replying by e-mail. ( ) |
Where is Bluetooth installation coming from and how can I removeit from my old Debian box? Posted: 23 Feb 2010 06:29 AM PST In article <4b8df04a$0$4975$newshosting.com>, jull43 @tampabay.rr.com says... None. That's what I've been trying to say... D |
Posted: 16 Feb 2010 11:46 PM PST balou wrote: Modern URL?. AFAIR, the only onlne HOWTO references all commented about unreliability. OTOH, these things are usually just cobble togethers with various hardware. |
Getting *all* local mail to be sent to the smarthost relay Posted: 16 Feb 2010 06:52 AM PST On Feb 17, 7:51am, The Natural Philosopher <invalid> wrote: You're referring to the "dpkg-reconfigure" script, or whatever it's called? That's not in the RHEL RPM's. I just had a gentle talk with the configure files in /etc/exim/ on RHEL 5. It works, but it's not well doented. I had to go hunting in the on-line manuals for details on setting up a smarthost, which involves replacing a stanza about DNS handling. Right. Exim Doesn't Do That(tm). There's a table in /etc/exim/ where you need to put such aliases instead. |
Posted: 15 Feb 2010 02:42 PM PST On Mon, 15 Feb 2010 17:42:35 -0500, Ranman <com> wrote: The dos mbr is limited to a total disk size (not just partition) of 2TB. Trying to create/write past 2TB will result in wrap around, causing corruption. See http://en.wikipedia.org/wiki/GUID_Partition_Table From "man fdisk" ... fdisk doesn't understand GUID Partition Table (GPT) and it is not designed for large partitions. In particular case use more advanced GNU parted(8). While your kernel appears to have guid support enabled, you have to use partitioning software that also supports it, and you have to select guid as the partition table type (aka gpt). With parted, you would use "mklabel gpt", followed by the normal partition creation. Regards, Dave Hodgins -- Change nomail.afraid.org to ody.ca to reply by email. (nomail.afraid.org has been set up specifically for use in usenet. Feel free to use it yourself.) |
Can my ubuntu desktop access web via wirelessrouter Posted: 14 Feb 2010 04:42 AM PST On Feb 14, 10:18pm, Nico Kadel-Garcia <com> wrote: thanx for that |
Newbie looking for advice on Linux distribution Posted: 13 Feb 2010 06:57 AM PST On Mon, 15 Feb 2010 11:33:42 +0000, Alan Mackenzie wrote: I have tried various distros over the years, and I always gave up due to various problems, notably hardware recognition issues. Last week I began looking at Ubuntu netbook remix. I downloaded it and installed it on my netbook. It picked up all my hardware and booted into a fully functional system, with only minor tweaking. This week, all my computers are running ubuntu. I have 7 machines here, and it installed on all of them with only one problem. My main laptop had the harddrive trashed by the installer. Luckily it was all backed up, but my Vista 64 was trashed. I was fuming, but I decided to continue installing Ubuntu. I got it installed and it is running flawlessly. Of course, nothing is perfect. I am a musician, and I like to do some multitrack recording. THis is the one area where Linux is very weak. I had to leave my DAW machine untouched, so that I could continue to use my studio. I'm not saying I can't record audio on a linux box, just that it is much more limited than Windows. I could just be missing things, and I'm aware that there is always a learning curve with any new endeavor, but it seems that the capabilities just aren't there. We'll see..... |
Protecting the Windows using Linux Posted: 12 Feb 2010 11:36 AM PST On Sunday 14 February 2010 12:58 in comp.os.linux.setup, somebody identifying as TomB wrote... Follow-up header respected, albeit that I haven't been subscribed to that group anymore in a few years already. Xen is primarily suited for server deployment, albeit that it can be used with workstation set-ups as well - see her down The thing about Xen is that it's a bare metal hypervisor, so it doesn't run inside a host operating system. Everything running on top of Xen - including the management system - is a virtual machine. It's similar to how mainframes work, but with the difference that the operating system in direct control of the hypervisor on a mainframe is a specialized single-user system, while on Xen it must be either GNU/Linux, NetBSD, OpenBSD or (Open)Solaris, all of which are UNIX-style systems and thus multi-user. It is however advised, especially for server set-ups, not to have any users log into the management virtual machine, or perhaps, just one user, and have that user then use /su/ and/or /sudo/ to obtain root privileges. (Note with regard to the above: I always disable all direct root logins, both remote and local, on all of my machines, virtual or physical, and thus an unprivileged user account must then be used to log in directly, and /su/ from there on. This means that any cracker breaking into the system must instead of guessing only the root password now guess my user account's login, its password, and then the root password.) If you intend to run GNU/Linux-only guest systems (and as servers), then you might also want to look into OpenVZ and Vserver as an alternative solution. This is another kind of virtualization, at the operating system level, i.e. you then run multiple userspace "containers" (also called "zones") on top of a common kernel, with one userspace context being "the host", from which you can access all others. OpenVZ and Vserver are similar to eachother but there are some important differences. For instance, Vserver uses a copy-on-write system for the guests which is economic in diskspace, but OpenVZ has more possibilities and uses a more recent kernel - 2.6.18 for "stable" and 2.6.26 and 2.6.27 for "testing". Another operating system which offers this kind of virtualization would for instance be (Open)Solaris. Personally I would like to see the OpenVZ adopted into the upstream Linux kernel. It already has a now quite mature Xen support built-in (for both dom0 and domU) and it also offers KVM and lguest as additional virtualization technologies, but those are too much akin to the third party virtual machine monitor set-ups of VMWare (Workstation) and VirtualBox. Operating system level virtualization would be a nice complement to GNU/Linux, especially since OpenSolaris already offers it as well, and if I'm not mistaken, then FreeBSD also already had it at one stage (although I think they've removed it again now - I'm not sure.) In the event of a Xen set-up, each of the virtual machines runs a complete operating system, i.e. kernel plus userspace. So there's a little more RAM overhead than with OpenVZ or Vserver. Otherwise, Xen performs very well in comparison. With paravirtualization, performance of the guests is only some 1% or 2% slower than if they were running on the bare metal. Another advantage is that Xen can run different types of guest operating systems. You can even run Windows as an unprivileged guest on Xen, but only on the condition that your hardware has virtualization extensions, because Windows can obviously not be paravirtualized, since the code is not free. Microsoft did at one stage - during the development of Xen - supply a paravirtualized version of Windows XP, but this version was never licensed for retail; it was solely intended for testing by the Xen developers. Performance-wise, hardware virtualization is slower than paravirtualization, though. With hardware virtualization, part of the hardware the HW virtual machine sees is emulated by Xen, using the Qemu device manager. This emulated hardware is also not exactly "the latest and greatest", but at least it works reliably. Paravirtualization on the other hand is an approach in which the unprivileged guest operating system is "aware" that it is running virtualized. A paravirtual guest has a kernel which uses so-called "front-end" drivers, which are basically an abstraction layer that connects to the "real" back-end drivers running in the dom0 virtual machine. So there is no emulation involved, and all of the systems running on that one physical machine actually become one big multifunctional virtual machine operating system. Again, it's like a mainframe system. Xen also allows the sysadmin to tailor performance by configuring how many virtual CPUs each guest can use, and for performance-critical virtual machines, it is possible to assign one or multiple physical CPU cores to them, so that the other virtual machines cannot use those. There are also multiple scheduling options for shared physical CPUs. On the networking side of things, Xen defaults to bridging, but it is possible to use routing/NAT as well - scripts are supplied to easily set up whatever configuration you prefer. Xen also supports isolating certain hardware from the management virtual machine (dom0) so that this hardware can be directly accessed by one of the unprivileged virtual machines (domU), which is then considered a "secondary driver domain"; this is again often applied in (Open)Solaris, even with Sun's own bare metal hypervisor - I forgot what it's called. In other words, if you're running a virtual machine which needs a lot of network bandwidth and your physical machine has two NICs, then you can choose to hide one NIC from dom0 and have this particular virtual machine access the second NIC directly with a regular driver and without having to use the bridging or routing via dom0's NIC. The machine I am currently working on - i.e. setting up; I'm not talking of the machine I am typing this from - is going to be running Xen with multiple paravirtualized Gentoo GNU/Linux virtual machines - the dom0 plus two domUs. One of the domUs will have direct access to a limited set of hardware - i.e. a dedicated PCIe video adapter card, the on-board sound chip and all USB hubs - which will then of course be hidden from dom0, and the second domU will be running an OpenVZ kernel with multiple "zones", installed as "headless servers". Virtualization on top of virtualization, and all of it is Free & Open Source Software. ;-) (Okay, that last line was specifically intended for COLA. :p) -- *Aragorn* (registered GNU/Linux user #223157) |
Intel Graphics Under Debian Lenny (Blank Screen) Posted: 10 Feb 2010 06:35 AM PST There are two intel drivers for X, i810 and intel (the newer one). Check out, which one Ubuntu and Lenny are using. -- flori Vim-Hilfe auf Deutsch http://www.florianrehnisch.de/vimhelp/ WARNING: PRGRAMMING BUG IN E2FSCK! OR SOME BONEHEAD (YOU) IS CHECKING A MOUNTED (LIVE) FILESYSTEM. |
cannot change current block/inode allocation Posted: 08 Feb 2010 10:05 AM PST Bill Mar wrote: I tried this , but still i am getting this zodiac@zodioc:~$ sudo su root@zodioc:/home/zodiac# edquota -u testuser edquota: WARNING - /dev/sdb5: cannot change current block allocation edquota: WARNING - /dev/sdb5: cannot change current inode allocation |
Posted: 08 Feb 2010 02:51 AM PST On Feb 8, 1:08pm, donottypethisbit.com (Mark Hobley) wrote: Linuxconf hasn't had an update in 5 years. The tools available for particular Linux distributions vary widely, and are often specific to specific features: "yumex" for yum software updates on RHEL, "system- config-network" for network setups on RHEL, "YaST" for all sorts of configurations on SuSE, "lilac" on various Fedora and Debian based platforms for Nagios, etc. Frankly, I prefer Webmin wherever feasible: it's a better interface and more robust, by far, than many of the other GUI's. But its completeness relies on people writing good components to manage parts of /etc/, and creating *those* is often not part of the author's plans. CUPS, for example, is just nasty to configure graphically due to missing access to core components, such as the text->Postscript page size settings. |
Posted: 07 Feb 2010 01:06 PM PST On Feb 10, 12:45pm, Peter Köhlmann <de> wrote: That's their problem. Well, that and SuSE. (Its primary configuration tool, YaST is really, really, really not your friend for any remotely sophisticated configuration needs due to a number of pervasive flaws in mishandling both kernel and package configurations, especially the "there can be only one kernel" and the 3rd-party software installations such as NVidia. And don't get me going on how foolish it is to publish a kernel SRPM where the patch files are hidden inside tarballs, to be plucked and chosen among by shell scripts. Errrggghhh!) Manuel actually asked an interesting question, the sort that's worth trying to answer and educate with even if it seems foolish to someone already more familiar with how these things work. For example: optimizing the "initrd", which is not actually a script but a compressed, mountable file system for old reasons involving bootstrapping from 1.44 MByte floppies, is an interesting way to try to speed things up. Optimizing the init scripts, which is a different matter, is actually something people are actively pursuing, and it's one of the lengthier parts of the boot procedure. Optimizing the BIOS itself is also possible, but takes considerable courage and some knowledge to do. (LinuxBIOS isn't broadly supported yet, but it's awfully cool if you can make it work.) Manuel, I'm glad you're more clear now on the difference between grub, the boot loader, and the init processes. They're tricky: they're shell scripts, fitting a very standardized layout to ease management, but they really *weren't* written for speed. For example, throwing out all that nice sanity checking of config files and testing whether they exist is going to slow things down. And the various funky bits designed to make the printouts look pretty are a complete waste of most user's time: We don't care of the output is "OK" or a splufty, pretty green OK printing subroutine, but we needn't waste our time loading up shell function files if we just print "OK". |
Posted: 06 Feb 2010 12:53 PM PST On Sat, 2010-02-06 at 12:53 -0800, syd_p wrote: I don't have a Bladesystem H yes... but in general the blade manufacturer provides a virtual CDROM/DVD capability that allows you to connect your "local" DVD from a the remote control point running the enclosure software and use that for the blade(s). So... you effectively get a DVD drive on your blades that comes off your desktop (if you will). Of course, once you do one machine, you could always clone it to the other blades... lots of options once you do one. |
Posted: 03 Feb 2010 03:32 AM PST On Feb 9, 4:36am, kaari <com> wrote: I'm no squid expert, but your request is short on what might be useful information. What OS is your proxy server? What version of Squid is it? And can you post the configuration file for it? Similarly, what OS is your web client? Which browser? And does this website work well *without* the proxy in the way? Last, what is the target website? |
You are subscribed to email updates from TextNData Forums - Linux To stop receiving these emails, you may unsubscribe now. | Email delivery powered by Google |
Google Inc., 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States |