| View previous topic :: View next topic |
| Author |
Message |
selig
Guru
Joined: 31 Jul 2005
Posts: 425
Location: Prague, Czech Republic
|
 Posted: Tue Aug 16, 2011 9:41 am Post subject: Block devices missing in /dev [SOLVED] |
 |
|
I have recently rebooted a server after updating to baselayout-2. However, block devices were completely missing after the reboot so I am pretty lucky SSH started at all and I could log in.
I created the device nodes manually and everything seems to work now. udev even created symlinks for /dev/md* devices.
I do not know what caused this. I am using hardened-sources-2.6.28-r9 and udev-151-r1. I know these are a bit old but could it have caused the missing device nodes? Currently, I have this in /dev:
| Code: |
block random tty10 tty36 tty61 usbdev6.1_ep81
bus root tty11 tty37 tty62 usbdev7.1
console rtc0 tty12 tty38 tty63 usbdev7.1_ep00
core sda tty13 tty39 tty7 usbdev7.1_ep81
cpu sda1 tty14 tty4 tty8 usbdev8.1
cpu_dma_latency sda2 tty15 tty40 tty9 usbdev8.1_ep00
disk sdb tty16 tty41 urandom usbdev8.1_ep81
fd sdb1 tty17 tty42 usbdev1.1 vcs
full sdb2 tty18 tty43 usbdev1.1_ep00 vcsa
grsec sdb3 tty19 tty44 usbdev1.1_ep81 vcsa1
hpet sdc tty2 tty45 usbdev2.1 vcsa12
char sdc1 tty20 tty46 usbdev2.1_ep00 vcsa2
initctl sdd tty21 tty47 usbdev2.1_ep81 vcsa3
input sdd1 tty22 tty48 usbdev3.1 vcsa4
kmsg sdd2 tty23 tty49 usbdev3.1_ep00 vcsa5
log sdd3 tty24 tty5 usbdev3.1_ep81 vcsa6
mcelog sg0 tty25 tty50 usbdev4.1 vcs1
md sg1 tty26 tty51 usbdev4.1_ep00 vcs12
md0 sg2 tty27 tty52 usbdev4.1_ep81 vcs2
md1 sg3 tty28 tty53 usbdev5.1 vcs3
md8 sg4 tty29 tty54 usbdev5.1_ep00 vcs4
mem shm tty3 tty55 usbdev5.1_ep81 vcs5
network_latency stderr tty30 tty56 usbdev5.2 vcs6
network_throughput stdin tty31 tty57 usbdev5.2_ep00 zero
null stdout tty32 tty58 usbdev5.2_ep81
port tty tty33 tty59 usbdev5.2_ep82
ptmx tty0 tty34 tty6 usbdev6.1
pts tty1 tty35 tty60 usbdev6.1_ep00
|
I created the md* and sd* device nodes manually.
my rc-status:
| Code: |
# rc-status sysinit
Runlevel: sysinit
dmesg [ started ]
udev [ started ]
devfs [ started ]
# rc-status boot
Runlevel: boot
hwclock [ started ]
modules [ started ]
fsck [ started ]
root [ started ]
hostname [ started ]
mtab [ started ]
localmount [ started ]
procfs [ started ]
urandom [ started ]
sysctl [ started ]
bootmisc [ started ]
termencoding [ started ]
net.lo [ started ]
consolefont [ started ]
keymaps [ started ]
syslog-ng [ started ]
swap [ started ]
# rc-status default
Runlevel: default
net.eth0 [ started ]
net.eth1 [ started ]
svscan [ started ]
ntp-client [ stopped ]
ntpd [ crashed ]
courier-pop3d-ssl [ started ]
udev-postmount [ started ]
proftpd [ started ]
netmount [ started ]
mysql [ started ]
sshd [ crashed ]
apache2 [ started ]
clamd [ started ]
courier-imapd [ started ]
courier-imapd-ssl [ started ]
courier-pop3d [ started ]
smartd [ started ]
spamd [ started ]
vixie-cron [ started ]
local [ started ]
|
"crashed" ntpd and ssh are bogus messages because of missing .pid (they started before I was able to mount /var)
| Code: |
# LANG=C df -h
Filesystem Size Used Avail Use% Mounted on
rootfs 9.2G 6.9G 2.3G 76% /
/dev/root 9.2G 6.9G 2.3G 76% /
rc-svcdir 1.0M 92K 932K 9% /lib64/rc/init.d
udev 10M 60K 10M 1% /dev
none 5.9G 0 5.9G 0% /dev/shm
temp 768M 1.2M 767M 1% /tmp
/dev/md1 915G 557G 358G 61% /var
|
Root is on /dev/md0. The system was able to boot because it has kernel-level raid autoconfiguration but I had to mount /dev/md1 manually because the device node had been missing until I created it...
What should I do/examine? I do not want to update the kernel and udev and reboot the system again just to find it in the same unusable state.
Last edited by selig on Sun Aug 28, 2011 10:02 pm; edited 1 time in total |
|
| Back to top |
|
 |
redagadir
n00b
Joined: 06 Aug 2011
Posts: 29
|
| Posted: Sat Aug 20, 2011 1:34 am Post subject: |
|
|
most probably udev didn't start, or if the problem is only related to /dev/md*, this means mdadm didn't do the scan at startup.
could you check with both?
Last edited by redagadir on Wed Dec 21, 2011 9:12 am; edited 1 time in total |
|
| Back to top |
|
|
selig
Guru
Joined: 31 Jul 2005
Posts: 425
Location: Prague, Czech Republic
|
| Posted: Tue Aug 23, 2011 6:37 pm Post subject: |
|
|
| Well, udev is running fine. And those raid arrays are automatically started during the kernel boot process (root is on one of them). Maybe it is really because of the old versions? I will have to reboot the server in a couple of days so I guess I will see what happens after I upgrade kernel and udev. |
|
| Back to top |
|
|
VoidMage
Watchman
Joined: 14 Oct 2006
Posts: 6196
|
| Posted: Tue Aug 23, 2011 7:55 pm Post subject: |
|
|
Well, in the migration guide, the need to add lvm/mdraid to boot, if you use them is mentioned.
Also, the changes in kernel module loading might be relevant.
But udev might just be pretending to run - there's i.e. CONFIG_SYSFS_DEPRECATED issue. |
|
| Back to top |
|
|
selig
Guru
Joined: 31 Jul 2005
Posts: 425
Location: Prague, Czech Republic
|
| Posted: Sun Aug 28, 2011 10:02 pm Post subject: |
|
|
| Updating udev and kernel to recent versions and making sure udev does not complain about any kernel options helped. |
|
| Back to top |
|
|
dtlgc
n00b
Joined: 27 Jul 2004
Posts: 60
Location: Texas
|
| Posted: Sun Sep 25, 2011 4:55 pm Post subject: |
|
|
Did you find a solution to this?
Having similar problem...
fdisk -l returns nothing....
looking in dev all devices are listed
>>>>>>>>>>>>
rebuilt kernel from this advice:
# mdadm --detail --scan >> /etc/mdadm.conf
# genkernel --mdadm --mdadm-config=/etc/mdadm.conf ramdisk
from here: https://forums.gentoo.org/viewtopic-t-881795-start-0.html
can now see devices with fdisk
>>>>>>>>>>>> |
|
| Back to top |
|
|
selig
Guru
Joined: 31 Jul 2005
Posts: 425
Location: Prague, Czech Republic
|
| Posted: Wed Sep 28, 2011 10:21 pm Post subject: |
|
|
| It seemed to be an issue with a deprecated kernel configuration option being turned on - and because of that, udev (which I had updated) did not work correctly. I turned all deprecated options off, updated the kernel and it started working again. However, I am not using genkernel, I am using hardened-sources. I can post my functional .config if it helps you. |
|
| Back to top |
|
|
dtlgc
n00b
Joined: 27 Jul 2004
Posts: 60
Location: Texas
|
| Posted: Fri Sep 30, 2011 3:03 am Post subject: |
|
|
| That .config might be nice to review - I am also using hardened-sources |
|
| Back to top |
|
|
selig
Guru
Joined: 31 Jul 2005
Posts: 425
Location: Prague, Czech Republic
|
|
| Back to top |
|
|
|
Kernel & Hardware
