I am trying to make gentoo boot faster and in my searching I came across this article http://lwn.net/Articles/299483/. One of the things they did was revert to a "persistent, old-school /dev directory so that boot doesn't depend on udev". I would like to know how to do this, but when I searched for disabling udev all the forums say that its a bad idea. All I want to know is how to make a static /dev directory.
View 4 RepliesI've recently updated my Slackware 13.1 system to the Slackware current. Although I have created my "initrd" image without specifying the "-u" option to "mkinitrd" it still starts up "udev".
That is causing me some difficulty because I am using "dmraid" to detect my RAID arrays. I had created my own device names such as "/dev/sdr2" for my root partition. With 13.1 I had no problem, since "udev" was not started by the "initrd" unless the "-u" option was provided. The current version seems to start up "udev" even without that option.
Is there a way to disable "udev" in the "initrd", or is there a way to specify custom "udev" rules for an "initrd"? I tried placing a "10-local.rules" file in the "etc/udev/rules.d" directory of the "initrd-tree" but that file had no effect on the device names generated by "udev" during the "initrd".
Here is my script that creates the "initrd".
Code:
ROOTDEVNAME="/dev/sdr2"# Name of root device
LINUXVER="2.6.35.7-smp"# Linux modules version
CLIBVER="2.12.1"# C library version
ROOTDIR="/boot/initrd-tree"# Location of root filesystm
[code]....
It will be helpful for me to understand "udev" issues related to an "initrd" because I will eventually try to use "mdadm" instead of "dmraid". So far I have only been able to get my system to boot from the RAID array using "dmraid" and I often run into new problems when I update Linux. Still, Slackware has proven to have the best support for booting from my RAID array because of the user community, documentation and flexibility.
Is there a way to make the 'loading linux...' Bios Data check faster or disabled? It is taking 3 times longer then the actual boot.
View 2 Replies View RelatedI was unsatisfied with the 40second boot time of lucid and was searching for a solution for a while but didn't find anything yet. But today I found a way to boot 10seconds quicker.Lucid is installed here as suggested by the installer:
Primary rootpartition (/dev/sda1)
Logical partition (/dev/sda4)
swap (/dev/sda5)
[code]....
I am using Ubuntu Maverick on my Eeepc 701, and everything is working quite nicely. Since I only have 4GB SSD drive, my setup is that the LiveCD is booted from the SSD, while my 'casper-rw' partition is on a 4GB SD card. I have 2GB of ram and do not use swap. I am wondering if there is a way to somehow cache the results of the hardware probe and configuration, and insert the cache into LiveCD by remastering it. The idea is that less time will be needed to boot, since everything that was found from the first boot was saved.
Of course this would mean that particular modified LiveCD wouldn't reliably boot on anything other than my system (or one like it), but seeing as how my hardware won't change in the future, it isn't a problem.
Is something like this possible? I'm not afraid of recompiling a kernel or rebuilding an initramfs if needed. A possible alternative idea to accomplish this would be to boot up the vanilla LiveCD as normal, configure a swapfile, hibernate, then inject this swapfile into the LiveCD image.
This way, every boot would automatically just be resuming from hibernation. This could potientally mess with the casper-rw partition, but that is something I would worry about later (and I am not opposed to just eliminating the casper-rw partition altogether and running off ram each boot).
The other problem might be that the swapfile would probably need to be the size of my ram (2GB). Chances are, after a fresh boot my ram will be mostly empty, therefore would it be possible to compress the swapfile (like swapfile.gz)? I will be looking into either alternative, but I was hoping to get some opinions / ideas on how to accomplish this (or whether it has already been done!)
[Code]..
Would be possible to have a dual boot with Ubuntu only loading the things necessary for command line use of emacs?
I thought it might be good to be able to take notes for class on a quickly booting command line. (It has the added benefit of being less distracting)
Does it work like that? Could a pared down command-line only install be sufficiently faster on boot-up? Might another flavour of Linux be more suited to the task?
see [URL] I am still somewhat happy with my netbook, although I have a Lenovo X61 from work (with XP) that I just love and since the netbook was purchase for travel, I don't use it as much as I use the x61. But I am preparing for a trip soon and have been trying to get the netbook to boot faster. Default Ubuntu 9.04 was taking about 1m 40s from off to a usable web site. Many tweaks later using the default hardware still, I get boots on average of 1m 10s or so.
I have tried fluxbox and xfce4, but shave maybe 10s off on a good boot, but am then in a much more limited DE, so I don't see the 10s being a good reason to limit myself. Using bootchart, it reports a boot time of ~30 so I really do believe it is gnome/X that is taking the longest time. Anyway...I finally broke down and got an ssd drive from Newegg [URL].
I dd'd the old hdd to the new ssd and guess what? Absolutely no change what so ever to the boot times. The ssd drive (although silent) has not gained me any speed at boot up. My average boot times are still 1m+. Also, I ran a battery test last night (wrote time out to a file every minute and ran totem none stop until the battery died) and I got 2 hours of life out of the 3 cell battery. This is about the same battery life I was getting with the traditional hdd.
In conclusion: As it stands now, the ssd drive has gained me nothing but silence, but at the cost of 100gigs (hdd = 160gig vs ssd = 60g)... no improvement to battery life no improvement to boot times The system does *seem* a little more responsive when the OS is finally up, but again, at what cost?
I have an Ubuntu Server 10.10 install on a VIA C7 chip. It seems to pause for about 5 min on a USB component.
It is a fresh install, here is a snip from my messages log code...
My KB is USB, so I would rather not disable it if at all possible. The next time I can take it down, I will test disabling it though...
How to make the system boot faster by removing the idle time between 5s to 10s? bootchart attached. It is Ubuntu10.04LTS by the way. One more hint, the screen black out for ~4s after "Begin: Running /scripts/init-bottom... Done." I don't know what is going on during that 4s, but my best guess is there is a way we can get rid of it.
Bootchart can be found here:
my ubuntu wait about 25 seconds on bootdmesg:
Code:
[ 0.815389] udev[90]: starting version 163
[ 0.904058] sky2: driver version 1.28
[code]....
During the boot process the machine (Fujitsu Celsius M470) hangs about 4 min at udev:loading drivers. After that it continues and I can work with the OpenSuse 11.2 system without problems.I activated the debug log in /etc/udev/udev.confthen I see that it is doing a lot in that time, at the end I see the message:udevadm settle timeout queue contains: a long list of pci/usb entries (no entries in any log for this)
in the logfile I find:udevd-work[1071]: '/sbin/modprobe' (stderr) 'FATAL: Error inserting ipmi_si (/lib/modules/2.6.31.12-0.1-desktop/kernel/drivers/char/ipmi/ipmi_si.ko): No such device'there is no ipmi. how can I deactivate the loading of this module?
udevd-work[412]: '/sbin/modprobe' (stderr) 'FATAL: Module input:b0019v0000p0001e0000_e0,1,k74,ramlsfw not found.'
Mar 17 11:35:36 mira udevd-work[418]: '/sbin/modprobe' (stderr) 'FATAL: Module
[code]....
In Ubuntu 910 on my Asus EEE-PC1005HA, I notice that during boot, when the screen is dark between the plain Ubuntu logo and the animated purple screen, there is a brief message:udev[488] cannot read file (rest of message goes by too fast for me)This doesn't seem to hurt anything. My system has been working just fine for a long time. I don't know how long the message has been there. Perhaps it merely lengthens boot time? Should I care? If I should care, where might the message be logged (so that I don't have to photograph the screen)?
View 5 Replies View RelatedI'm writing for my brother who was running Karmic on a Microtel desktop.He'd been having trouble with blank cds (data cds and dvds didn't cause this problem) crashing his box so he uninstalled and then installed udev package.Now when he tries to boot the box goes into a memory test and then reboots into the memory test over and over. He tried to check the installed kernels to choose one that might boot but there are none listed (he pressed "esc" to see the list-it's empty).He tried booting from a live cd and that won't work either. The cd drive spins but nothing else.
View 1 Replies View RelatedThis morning I had the courage to run an apt-get autoremove which took some 275 packages off my hard drive, and now I am experiencing a delay of about 40 seconds at boot, after grub, before the plymouth splash appears. The cursor blinks on a black screen while the hard drive churns away. Finally two error messages appear too quickly to be read, and then the bootsplash kicks in. I can find the instant in the dmesg where the delay happens, but can't locate the cause. Here's what my dmesg looks like:
[Code]....
The laptop is an Acer Timeline X 3820TG, with the dual GPU "switchable graphics." These dual graphics cards have given me enough trouble in the past that I wouldn't be surprised if they were the problem. But the hard drive action sounds like a 'fsck,' and seems to be contemporaneous with the dmesg notice that the root partition is mounted.
Incidentally, my boot wasn't all that fast before; I would not be surprised if this delay was preexisting, but used to happen after the plymouth boot screen was already on screen. Still, if I can get rid of this one ugly delay, I can have a fast (c. 10 secs) boot time.
I've been having problems with start_udev on my machine post kernel 2.6.18-162 on my CentOS machine. A previous install would work on the old kernelut not the newer ones. I recently did a completely fresh install of CentOS 5.5o see if maybe there was a orruption in my install when it upgraded,t I still have the problem of udev hanging on startup.Using the rescue option on the disk, I've managed to track the problem down to the 50-udev.rules file. Through trial and error, I moved all the .rules files out of the rules.d directory, and added them back one at a time to see which one(s) caused the system to hang, and which ones it carried on booting as normal with. The only one that causes it to hang is the 50-udev.rules.
From what I can tell, this rules file is responsible for letting udev check various pieces of hardware. I think that it is failing whent runs modprobe on a pci device address.Only problem is I don't understand the rules file syntax, so I don't know if I can simply comment out or change a line in this file to skip the device that its looking for.Can anybody help me track down the specific device/line that is doing this?(I've kept the file out of the folder for now and my system appears to continue operating, but I get the occasional problem that for all I know could be due to udev not having a complete start up).
I have done a fresh text-only installation of Fedora 10 on a Dell Dimension E521 for the purpose of setting up a server. After installation I ran yum update to bring the system up to date. After the update, I rebooted the computer, but the boot process froze. I recycled power and pressed "I" after the Dell BIOS screen and the GRUB bootloader appeared. I selected the most current version, edited the kernel line by deleting "rhgb quiet" and replacing it with "3." After making this change, I continued with the boot and the computer stopped at "Starting udev:" I have two fedora 10 revisions showing in the GRUB bootloader, the original installation and the update after running yum.
I repeatedly tried rebooting both versions and, after about 30 attempts, the computer finished booting and got me to the command prompt. Reading through the forums indicated there might be some issue with my nVidia GeForce 6150 onboard video and fedora 10. So, when I got to the command prompt, I followed the instructions in the forums [URL] to load the rpmfusion drivers. This appeared to be successful and when I looked at /etc/X11/xorg.conf it appeared to be correct for the new nVidia drivers. After loading these new drivers I tried rebooting. Unfortunately, I have been trying now over and over to get back to the command prompt, but simply can't get past "Starting udev."
i'm heaving the following message at boot time: "Starting udev: udevd[114]: unknown key 'DEVTYPE' in /etc/udev/rules.d/70-cups-libusb.rules:6", i guess it's something about usb printing support but i am not sure, and i don't know how to fix it, does any of you guys know what this exactly means and how to fix it
View 6 Replies View RelatedI'm trying to configure gpsd 2.96 to start automatically from udev rule on a Slack 13.37 box.1. I've compiled and installed gpsd from sources and made sure it starts manually.2. I've copied the /lib/udev/gpsd.hotplug and /lib/udev/gpsd.hotplug.wrapper scripts in their places from the source tree and made them executable.3. I've copied the gpsd.rules file from sources into /etc/udev/rules.d4. I renamed it 99-persistent-gpsd.rules to run late in the bootup process.5. I've copied the /etc/default/gpsd file from sources and made sure it has the right settings inside.
Now, for the results. If I plug the gps usb dongle in while the system runs, it starts gpsd if it is not started, and it connects to it just as it should. But if I start the system with the dongle in, gpsd doesn't get started during boot. I can't find any relevant message in /var/log/syslog or /var/log/messages for boot time. There is stuff there from when I plug the dongle in while the system is running though. It's like udev ignores the rules for it at boot time.Is there something in the Slackware boot scripts that would prevent running those scriptssd.hotplug.wrapper which in turn runs /lib/udev/gpsd.hotplug which is a Python script)?Here is the contents of 99-persistent-gpsd.rules (ignore the comments referring to Debian, it was meant for a Debian box). My usb gps adapter is the first one - the Prolific chipset one:
Code:
# udev rules for gpsd
# $Id: gpsd.rules 5861 2009-08-03 13:41:01Z bzed $
[code]....
I have just installed slack-13.1 on an acer aspire one netbook from a usb stick by booting the install kernel with noudev. I was able to do a complete install including lilo but upon rebooting my boot hangs when encountering my webcam.I see enough info to note the id as 0c45:62c0. This is a microdia webcam which I may be able to do something about later but in the first instance I'd like to be able to boot my system. The bios is very basic and there is no way to disable devices.
View 14 Replies View RelatedHave just installed 5.4 64 bit on an AMD64 x2 system with 4 GB ram running ESXi4.Text based install went fine, but on reboot starting up it gets as far as 'Starting udev' and just hangs. Checking the performance in ESXi it appears to be using 100% cpu.I have left it for half an hour and it does not progress and the only thing to do is to power cycle the VM.I have searched and found a few suggestions for kernel parameters but they did not make any difference. I can't even get in to a command line as it doesn't boot up far enough.I have reinstalled it several times and also checked the MD5 of the downloaded file and all appears Ok.
View 2 Replies View RelatedI've done an upgrade from lenny to squeeze.Now, I'm getting a lot of udev warning messages at the boot time.Knows someone howto remove them?
View 6 Replies View RelatedDebian testing: Kernel: Linux 2.6.32-5-amd64I can now only boot into safemode because I cannot get these 4 packages to upgrade:
udev
media-player-info
xserver-xorg-core
[code]....
I'm hoping someone knows about this one... I'm running the latest CentOS 5.4 with kernel 2.6.18-164.15.1.el5 (x86_64).When I boot the machine, it gets to the udev starting bit, hangs for like 5 minutes, then prints a message "Failed, will continue in the background." Then it boots OK after that.I tried booting again with the kernel option (from grub) "udevdebug", and what I saw when it tried again was a million messages saying it was waiting for "/sbin/pam_console_apply" to return, but I guess it wasn't returning... ;) Again, after 5 minutes, it gave up and finished booting.Now, this host is an LDAP client.
I figured that may have something to do with it as it is likely that pam_console_apply tries to make an LDAP lookup, which is wrong, because networking hasn't even started yet. If I disable LDAP (by removing ldap lookups in nsswitch.conf), I get no pam_console_apply errors from udev and it boots quickly. But that's a bummer, I need LDAP on this box, and I don't want my boot time to be 7-8 minutes. ;)Presumably before, when LDAP was enabled and it waited 5 minutes and then notified me that it will "continue in the background", that it was eventually successful after networking started. LDAP otherwise works fine on this box, just like all the other servers we have.This is new behavior, I've not seen it with CentOS 5.3 and below. Has anyone seen this? Any hints on what I can do to avoid it? It seems like a pam bug or something, but I don't know for sure.
Fedora 12 randomly crashes after a fresh install.Everything is left at its default during the installation except I install KDE and not Gnome.When I try to boot I hit esc at the splash screen to see where it hangs, but it never hangs at the same spot twice. Sometimes it hangs and sometimes it restarts. Its crashed everywhere from starting udev to the login screen.
I've been using kubuntu for the past year because I have the same problem with F10 and F11.
I dual boot with Windows 7 for games.
System specs:
Motherboard: MSI P45 Platinum
Processor: Intel Core 2 Duo E8500
Video Card: Radeon HD4850
Memory: 4GB (2x2GB) Patriot
HDD: Main: 250GB SATA Western Digital
code....
I just performed an update thru the Update Manager, & on re-boot got this message: ( Ubuntu 10.04.1 LTS, kernel 2.6.32-24-386 ) udevadm trigger is not permitted while udev is unconfigured. Gave up waiting for root device. Common problems :
- Boot args ( cat /proc/cmdline )
- Check root delay= ( did the system wait long enough? )
- Check root= ( did the system wait for the right device? )
- missing modules ( cat /proc/modules; ls /dev )
ALERT! /dev/disk/by-uuid/df0200e3-e6e9-439a-922f-100d92af0c58 does not exist. Dropping to a shell!. BusyBox v1.13.3 (Ubuntu 1:1.13.3-1ubuntu11) built in shell (ash) (initramfs)_ Info: I can boot into older version: Ubuntu 10.04.1 LTS, kernel 2.6.32-23-386. The Update Manager does not seem to want to show any updates that are available.
I upgraded to the last ubuntu last week and since I get the error: "udevadm trigger is not permitted while udev is unconfigured."I followed different posts on the forums but none helped: [URL]I did the sequence:
"1. Boot liveCD
2. "sudo fdisk -l" to find your boot disk, in my case it is /dev/sda5.
3. "sudo mkdir /media/newroot"
[code]....
I switched today to slackware-current on one of my desktops to play with it and ran directly into a problem.
Since ages my lilo.conf has two entries for slackware. One for runlevel 3 and one for runlevel 4.
Code:
Since the upgrade this is no more possible because I get a kernel panic as soon as udevadm trigger is called. The stack says something about an unknown boot option. Because that i removed the append lines from my lilo.conf and i was able to boot the system. The crash happens when udev is called from within the ramdisk and afterwards. I tried both.
My question is now. Is this a bug in udev or expected? I have this setup since at least 5 years and had never problems with that. What do I have to do to be able to select the runlevel at boot time?
This on a Vostro 1220 Laptop w/ Intel 5300 wireless:
A.I have long boot up time.I think it's because of the eth0 network search which I don't use.I have an intel wireless 5300 card running.How can I speed up the boot time, i.e. disable or change the eth0 at boot, the searching?
B:When I restart or shutdown, the screen flashes repeatedly and gets some garbled colors along the top before finally rebooting looks like windows ME or something).This vostro has an intel x4500HD vid chipset in it.
C.How do I get into gnome configuration editor to turn on Metacity compositing? Alt-F2 and run gconf-editor doesn't do it. I don't do compiz, but need compositing.
D.I need to install Chromium Browser as it sync my bookmarks.I have RPMforge enabled btw also...how can I do that? I.e. rpm repo for chromium?
This will help me get off to a running start so I can get up to speed on CentOS..
When booting Fedora 11, my system hangs for a very long time on starting udev. Sometimes I get an I/O error. However, my hardware is fine. I do eventually get in to the system.
View 7 Replies View RelatedOn a Sun Ultra10 333MHz, 512M, 9gB HDD. Booting Fedora-9. Silo v1.4.14 into kernel 2.6.27 64bit (vmlinuz-2.6.27.12-78.2.9fc9.sparc64). This is a brand-new installation.Although it's running on a Sparc it makes v.litte difference so far as this bootprocess, teh way Linux runs, where everything is - is concerned. That's why I've cross posted this query here.Booting merrily, in the interactive startup section just past "Starting udev", "Setting hostname", at "Checking filesystems" I get:
/: clean, 155284/557056 files, 920932/2225412 blocks
fsck.ext2: Device or resource busy while trying to open /dev/sda
Filesystem mounted or opened exclusively by another program? [FAILED]
*** An error occured during the filesystem check
*** Dropping you to a shell, the system will reboot
*** when you leave the shell
*** Warning --SELinux is active