+++ This bug was initially created as a clone of Bug #539273 +++ Description of problem: I have a VIA VT8237 SATA RAID controller, RAID-0 on two 250Gb disks. Has been working fine on F11. I upgraded from the commandline F11->F12, after boot I get 'no root device, sleeping forever'. Rebooting and selecting an older F11 kernel boots OK. Version-Release number of selected component (if applicable): old kernel boots OK: 2.6.30.9-90.fc11.x86_64 new kernel boot fails: 2.6.31.5-127.fc12.x86_64 Also, from the F12 install CD Anaconda reports that the disks must be initialized so no RAID appears not to be detected. How reproducible: Steps to Reproduce: 1. upgraded F11->F12 from commandline 2. reboot Actual results: boot halts 'no root device, sleeping forever' Expected results: boots Additional info: I read a similar bug id=522076, maybe related? Here's some commands: [root@katana giles]# parted /dev/mapper/via_djbccigdhe GNU Parted 1.9.0 Using /dev/mapper/via_djbccigdhe Welcome to GNU Parted! Type 'help' to view a list of commands. (parted) p Model: Linux device-mapper (striped) (dm) Disk /dev/mapper/via_djbccigdhe: 500GB Sector size (logical/physical): 512B/512B Partition Table: msdos Number Start End Size Type File system Flags 1 32.3kB 210MB 210MB primary ext3 boot 2 214MB 105GB 105GB extended 5 214MB 105GB 105GB logical lvm 3 105GB 315GB 210GB primary ext4 4 315GB 500GB 185GB primary ntfs [giles@katana ~]$ sudo cat /boot/grub/menu.lst # grub.conf generated by anaconda # # Note that you do not have to rerun grub after making changes to this file # NOTICE: You have a /boot partition. This means that # all kernel and initrd paths are relative to /boot/, eg. # root (hd0,0) # kernel /vmlinuz-version ro root=/dev/mapper/VolGroup-lv_root # initrd /initrd-version.img #boot=/dev/via_djbccigdhe default=2 timeout=15 splashimage=(hd0,0)/grub/splash.xpm.gz hiddenmenu title Fedora (2.6.31.5-127.fc12.x86_64) root (hd0,0) kernel /vmlinuz-2.6.31.5-127.fc12.x86_64 ro root=/dev/mapping/VolGroup-lv_root initrd /initramfs-2.6.31.5-127.fc12.x86_64.img title Fedora (2.6.31.5-122.fc12.x86_64) root (hd0,0) kernel /vmlinuz-2.6.31.5-122.fc12.x86_64 ro root=/dev/mapper/VolGroup-lv_root SYSFONT=latarcyrheb-sun16 LANG=en_US.UTF-8 KEYTABLE=uk initrd /initramfs-2.6.31.5-122.fc12.x86_64.img title Fedora-base (2.6.30.9-90.fc11.x86_64) root (hd0,0) kernel /vmlinuz-2.6.30.9-90.fc11.x86_64 ro root=/dev/mapper/VolGroup-lv_root initrd /initrd-2.6.30.9-90.fc11.x86_64.img title Other rootnoverify (hd0,3) chainloader +1 [root@katana giles]# blkid -o udev -p /dev/mapper/VolGroup-lv_root ID_FS_UUID=7da72cc3-e416-4c95-89eb-99d5934f95f0 ID_FS_UUID_ENC=7da72cc3-e416-4c95-89eb-99d5934f95f0 ID_FS_VERSION=1.0 ID_FS_TYPE=ext4 ID_FS_USAGE=filesystem [root@katana giles]# rpm -q dmraid dmraid-1.0.0.rc16-4.fc12.x86_64 [root@katana giles]# rpm -q dracut dracut-002-13.4.git8f397a9b.fc12.noarch [root@katana giles]# rpm -qf /usr/share/dracut/modules.d/* | sort -u dracut-002-13.4.git8f397a9b.fc12.noarch
*** Bug 539273 has been marked as a duplicate of this bug. ***
root=/dev/mapping/VolGroup-lv_root are you sure this is correct??? shouldn't it be "mapper" instead of "mapping" ??
(In reply to comment #2) > root=/dev/mapping/VolGroup-lv_root > > are you sure this is correct??? shouldn't it be "mapper" instead of "mapping" > ?? Yes, sorry, a late night typo! It was originally 'mapper'.
try to add "rd_NO_MDIMSM" for debugging add "rdinfo rdshell" or "rdinitdebug rdshell" you are then dropped to a shell and with rdinitdebug you can inspect "dmesg" messages and also /init.log. you might mount /boot by hand and copy over /init.log to attach it here.
OK, I added the debug command and dropped to the shell yesterday but wasn't sure where /boot was, I found /dev/sda1, but got this: mount -t ext3 /dev/sda1 /boot ext3-FS no journal found. There are no devices in /dev/mapper at this point only 'control'. No /dev/dm-x devices either.
ah, doh! your /boot is on the raid also... one thing you can do: 1. recreate the initramfs # dracut -f /boot/initramfs-2.6.31.5-127.fc12.x86_64.img 2.6.31.5-127.fc12.x86_64 2. boot with "rdshell" 3. get dropped to the shell and run # dmraid -l # dmraid -s 4. make a screenshot (camera, mobile phone) 5. attach the photo to this bugzilla
Created attachment 372538 [details] No block devices found
Pls see attachment 'no block devices found' Here's the output of dmraid -s when booting OK with old kernel: [giles@katana ~]$ sudo dmraid -s *** Active Set name : via_djbccigdhe size : 976794112 stride : 128 type : stripe status : ok subsets: 0 devs : 2 spares : 0
I've got the same problem trying to install FC12 x86_64 on a VT8237 BIOS RAID 0 (mirroring) with two 500Gb SATA disks: Anaconda doesn't recognize the HW RAID, allowing to perform the installation only on the two disks in a "standard" configuration (i.e. not-RAID). I can just add that I didn't face this problem during the installation of FC11 x86_64
ok, can you give me the output of # lsmod # dmsetup ls --tree from the boot with the old kernel
Created attachment 373057 [details] lsmod
dmsetup ls --tree via_djbccigdhep4 (253:4) └─via_djbccigdhe (253:0) ├─ (8:16) └─ (8:0) VolGroup-lv_swap (253:7) └─via_djbccigdhep5 (253:5) └─via_djbccigdhep2 (253:2) └─via_djbccigdhe (253:0) ├─ (8:16) └─ (8:0) via_djbccigdhep3 (253:3) └─via_djbccigdhe (253:0) ├─ (8:16) └─ (8:0) VolGroup-lv_root (253:6) └─via_djbccigdhep5 (253:5) └─via_djbccigdhep2 (253:2) └─via_djbccigdhe (253:0) ├─ (8:16) └─ (8:0) via_djbccigdhep1 (253:1) └─via_djbccigdhe (253:0) ├─ (8:16) └─ (8:0)
- boot with "rdshell" - get dropped to the shell and run # ls -l /dev/sd? # modprobe sata_via # ls -l /dev/sd? - make a screenshot (camera, mobile phone) - attach the photo to this bugzilla
Created attachment 373175 [details] modprobe
Created attachment 373183 [details] modules
Created attachment 373186 [details] after boot /lib/modules
Created attachment 373261 [details] lsmod from FC11 x86_64
Created attachment 373262 [details] dmsetup from FC11 x86_64
you created the initramfs with the wrong kernel argument :) please run (one line) # dracut -f /boot/initramfs-2.6.31.5-127.fc12.x86_64.img 2.6.31.5-127.fc12.x86_64 this might fix the problem. Or you have to rerun comment #13
# dracut -f \ /boot/initramfs-2.6.31.5-127.fc12.x86_64.img \ 2.6.31.5-127.fc12.x86_64
Hi Harald, Didn't we already do this on comment #6?
according to the screenshots with the missing modules, you did it wrong ( forgot the second argument )
OK definitely one line now: dracut -f /boot/initramfs-2.6.31.5-127.fc12.x86_64.img 2.6.31.5-127.fc12.x86_64 W: Possible missing firmware ql8100_fw.bin for module qla2xxx.ko W: Possible missing firmware ql2500_fw.bin for module qla2xxx.ko W: Possible missing firmware aic94xx-seq.fw for module aic94xx.ko Will now repeat comment #13 and attach sshots if there's still a problem...
Created attachment 373499 [details] dracut then boot into debug shell
Also tried rd_NO_MDIMSM kernel param and got this before the 'boot failed' message: "dracut: rd_MD_NO_MDIMSM no MD RAID for imsm/isw raids"
hmm, repeating comment #6 with the now correctly build initramfs from comment #23 would be cool. - boot with "rdshell" - get dropped to the shell and run # dmraid -l # dmraid -s
Created attachment 373787 [details] dmraid -l dmraid -s
looks great! should really boot! if not: - boot with "rdshell" - get dropped to the shell and run # blkid /dev/sda # blkid /dev/sdb # ls -l /dev/mapper # lvm vgchange -ay # ls -l /dev/mapper
and of course attach the screenshot, please
Created attachment 374038 [details] blkid /dev/sd
oh .. wow.. no output from blkid.. no need to reboot with rdshell.. just from the running system: # blkid -p /dev/sda # blkid -p /dev/sdb
no output from this either! # blkid -p /dev/sda # blkid -p /dev/sdb #
#blkid -p /dev/mapper/via_djbccigdhep5 /dev/mapper/via_djbccigdhep5: UUID="zsWDaI-r90B-medV-djja-dKWo-YkK7-FS7geu" VERSION="LVM2 001" TYPE="LVM2_member" USAGE="raid"
reassigning to util-linux-ng for further investigations. Karel, it seems like blkid does not report the drives as raid members.
I need last few blocks from the device. Please, from running system (you don't have to reboot with rdshell): # dd if=/dev/sda of=~/via.img bs=512 skip=$(($(blockdev --getsz /dev/sda)-10)) count=10 # gzip ~/via.img and send me the final ~/via.img.gz (or create an attachment to this BZ).
# dd if=/dev/sda of=~/via.img bs=512 skip=$(($(blockdev --getsz /dev/sda)-10)) count=10 10+0 records in 10+0 records out 5120 bytes (5.1 kB) copied, 0.000313716 s, 16.3 MB/s
Created attachment 374297 [details] via.img
util-linux-ng-2.16-10.3.fc12 has been submitted as an update for Fedora 12. http://admin.fedoraproject.org/updates/util-linux-ng-2.16-10.3.fc12
Giles, thanks for the test image. The problem should be fixed in util-linux-ng-2.16-10.3.fc12.
Bit confused, I haven't installed the new util-linux-ng yet. I enabled the test repo and updated thinking that the util-linux-ng update was in there. Machine booted OK (to cmnd line at least) with new kernel: 2.6.31.6-145.fc12.x86_64 ..but util-linux-ng still on 2.16-10.2.fc12 In dmesg dracut output showed something suspicious: dracut: Scanning for dmraid devices dracut: Found dmraid sets: dracut: via_djbccigdhe dracut: Activating via_djbccigdhe dracut: The dynamic shared library "libdmraid-events-via.so" could not be loaded: dracut: libdmraid-events-via.so: cannot open shared object file: No such file or directory dracut: RAID set "via_djbccigdhe" was activated dracut: Scanning devices dm-5 for LVM volume groups dracut: Reading all physical volumes. This may take a while... dracut: Found volume group "VolGroup" using metadata type lvm2 dracut: 2 logical volume(s) in volume group "VolGroup" now active
util-linux-ng-2.16-10.3.fc12 has been pushed to the Fedora 12 stable repository. If problems still persist, please make note of it in this bug report.
I'm sorry if this is out of the topic, but i don't know where else i can ask, so here it goes.. I have a FC11 x86_64 installed on a on a VT8237 BIOS RAID with two SATA disks and everything works just fine. Problem is, when I try to install FC12 from DVD on the same machine, Anaconda does not detect the BIOS RAID. Thank for any suggestion you may provide.
Hi Enrico Maybe do what I did, i.e. an online upgrade FC11->FC12 once the fixes are included in the main repo. You might have to wait longer to get an updated DVD.
Thanks Giles, I'll do that: can't wait for FC13..
(In reply to comment #41) > util-linux-ng-2.16-10.3.fc12 has been pushed to the Fedora 12 stable > repository. If problems still persist, please make note of it in this bug > report. Unfortunately util-linux-ng-2.16-10.3.fc12 does not fix the problem with me (I made up a DVD.iso with the updated package: during the installation anaconda still does not detect the VT 8327 BIOS RAID)