Opened 5 years ago

Closed 21 months ago

#633 closed defect (worksforme)

No filesystem could mount root during restore

Reported by: victorfeng Owned by: bruno
Priority: highest Milestone: 3.2.2
Component: mondo Version: 3.0.1
Severity: blocker Keywords:
Cc:

Description (last modified by bruno)

I am trying to restore a RedHat6.2 image created by mondorescue(3.0.1). During PXE network boot, I got following error:

List of all partitions:

No filesystem could mount root, tried: iso9660
kernel panic - not syncing: VFS: unable to mount root fs on unknown-block(1,0)
Pid: 1, comm: swapper Not tainted 2.6.32-220.el6.x86_64 #1
Call Trace:

I basically followed README.pxe.

It works fine with RedHat5.

During mondoarchive, I noticed "Disk /dev/mapper/vg_dcvops01-lv_root doesn't contain a valid partition table" poped up, which showed up on 'fdisk -l'.

[Redhat 6.2]# fdisk -l

Disk /dev/sda: 53.7 GB, 53687091200 bytes
255 heads, 63 sectors/track, 6527 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x0009758d

Device Boot Start End Blocks Id System
/dev/sda1 * 1 64 512000 83 Linux
Partition 1 does not end on cylinder boundary.
/dev/sda2 64 6528 51915776 8e Linux LVM

Disk /dev/mapper/vg_dcvops01-lv_root: 48.9 GB, 48930750464 bytes
255 heads, 63 sectors/track, 5948 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

Disk /dev/mapper/vg_dcvops01-lv_root doesn't contain a valid partition table


[Redhat 5]# fdisk -l

Disk /dev/sda: 214.7 GB, 214748364800 bytes
255 heads, 63 sectors/track, 26108 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sda1 * 1 65 522081 83 Linux
/dev/sda2 66 26108 209190397+ 8e Linux LVM

Victor Feng

Attachments (8)

mondoarchive.log (214.0 KB) - added by victorfeng 5 years ago.
mindi.log (133.5 KB) - added by victorfeng 5 years ago.
mondorestore_screenshot.doc (75.0 KB) - added by victorfeng 5 years ago.
cannot access mondorestore.log
mondorestore was not found.doc (97.5 KB) - added by victorfeng 5 years ago.
mondorestore was not found
No filesystem could mount root2.doc (75.0 KB) - added by victorfeng 5 years ago.
mondorestore was not found 2
No filesystem could mount root3.doc (98.0 KB) - added by victorfeng 5 years ago.
mondorestore was not found4.doc (114.0 KB) - added by victorfeng 5 years ago.
mondorestore command not found.doc (69.5 KB) - added by victorfeng 4 years ago.

Download all attachments as: .zip

Change History (31)

comment:1 Changed 5 years ago by bruno

Could you share the /var/log/mondoarchive.log and /var/log/mondorestore.log files so I can see better what is happenin in your env ?

Changed 5 years ago by victorfeng

Changed 5 years ago by victorfeng

Changed 5 years ago by victorfeng

cannot access mondorestore.log

comment:2 Changed 5 years ago by victorfeng

Hi Bruno,

The PXE restore hang from the very beginning. So there is no mondorestore.log. The screenshot shows where the restore hang.

Thank you Victor

comment:3 Changed 5 years ago by bruno

  • Description modified (diff)
  • Status changed from new to assigned

comment:4 Changed 5 years ago by bruno

Ok, so can you share your PXE conf file please, so I have an idea of how you call mondo at restore time at least.

comment:5 Changed 5 years ago by victorfeng

Have you tested MondoRescue? on RedHat6(6.2)? Did it work for you?

Here is the PXE conf which is recommended in README file of Mondo. It worked fine for RedHat? 5.

[root@prvkck01 pxelinux.cfg]# cat dcvops01 
default mondo
label mondo
  kernel img/dcvops01/vmlinuz
  append initrd=img/dcvops01/initrd.img load_ramdisk=1 prompt_ramdisk=0 ramdisk_size=131072 rw roo
t=/dev/ram iso_mode acpi=off apm=off devfs=nomount exec-shield=0 pxe prefix=dcvops01-20120621-arch
ive nfsmount=10.142.111.21:/mondo
[root@prvkck01 pxelinux.cfg]# 
Last edited 5 years ago by bruno (previous) (diff)

comment:6 Changed 5 years ago by bruno

Could you try increasing the ramdisk size, as it seems to not been recognized correctly, so maybe you lack memory.

Try:

ramdisk_size=1000000

comment:7 Changed 5 years ago by victorfeng

Changed ramdisk_size=1000000, and it does not make difference.

Victor

comment:8 Changed 5 years ago by bruno

To answer to a previous question yes, I did multiple RHEL6.x tests without particular issue. However, I remarked that you were using 3.0.1. Could you update to 3.0.2 or the beta of 3.0.3 instead ?

Same for mindi which is 2.1.1 to 2.1.3.

Also ensure you have mindi-busybox 1.18.5 as previous version also had an error related to what you mention (sorry didn't thought to that first).

Last edited 4 years ago by bruno (previous) (diff)

Changed 5 years ago by victorfeng

mondorestore was not found

comment:9 Changed 5 years ago by victorfeng

Hi Bruno,

I upgraded to 3.0.2. mindi to 2.1.3. mindi-busybox to 1.18.5 in the client box. During the PXE install, I ran into "The mondorestore command was not found on your backup media" error. Please see the attachment "mondorestore was not found".

Victor

comment:10 Changed 5 years ago by bruno

could you look upper in your screen, has there should a problem with the NFS mount IMO

Changed 5 years ago by victorfeng

mondorestore was not found 2

comment:11 Changed 5 years ago by victorfeng

I did not. But I did see Can't find CD's /mnt/cdrom/images/all.tar.gz.

Please see the attachment.

comment:12 Changed 5 years ago by bruno

When you increase the ramdisk, then you go further (it does make a difference). But not enough and mondorestore is not found as your NFS mount point is not mounted correctly. THat's this error tha would be interesting to see.

Changed 5 years ago by victorfeng

comment:13 Changed 5 years ago by victorfeng

I increased it to 5G, and got same error. I noticed error "udevd: specified group 'cdrom' unknown". Please see the attachment.

default mondo label mondo

kernel img/dcvops01/vmlinuz append initrd=img/dcvops01/initrd.img load_ramdisk=1 prompt_ramdisk=0 ramdisk_size=5000000 rw root=/dev/ram iso_mode acpi=off apm=off devfs=nomount exec-shield=0 pxe prefix=dcvop

s01-archive-20121025 nfsmount=10.142.111.21:/mondo

comment:14 Changed 5 years ago by bruno

You have 2 types of errors. One linked to the lack of memory, another onelinked to the fact you do not mount your NFS share (or at least that's what I guess with the info you gave).

With 5GB you don't see your kernel crashing don't you ? So find the NFS error on the screen (go back with Shuft PgUp? if needed) and fix what is not working if you can or report why it's not working at that moment.

udev prints tons of such error msgs which are harmless.

Changed 5 years ago by victorfeng

comment:15 Changed 5 years ago by victorfeng

I could not Shift PageUp? to the very begining. I could go up to about 10 pages. Attached is the page where could not go further. I did not see kernel crashing, nor NFS error.

comment:16 Changed 5 years ago by bruno

Also can you ensure that you NIC driver is indeed put in the backup ?

comment:17 Changed 5 years ago by victorfeng

How can I confirm that?

It is the same virtual machine running RHT5 and RHT6, and I used same command mondoarchive. When it is RHT5, restore succeeds. When it is RHT6, it does not.

Victor

comment:18 Changed 5 years ago by bruno

From your log you seem to be using an Intel e1000 NIC. The driver seems to be included in the backup:

Adding lib/modules/2.6.32-220.el6.x86_64/kernel/drivers/net/e1000/e1000.ko (292 KB) to the rootfs

Can you try to relaunch the start-nfs script manually in the console and report the errors seen there.

comment:19 Changed 4 years ago by bruno

Could you try again with the latest version of MondoRescue to see whether you stil lhave the issue ?

comment:20 Changed 4 years ago by victorfeng

I just downloaded the latest MondoRescue? 3.0.4 today. It does not work when I tried to restore. But the error is different:

"The mondorestore command was not found on your backup media"

Please the attached screenshot.

victor

Changed 4 years ago by victorfeng

comment:21 Changed 4 years ago by victorfeng

Actually, the error message is similar.

comment:22 Changed 2 years ago by bruno

  • Milestone changed from 3.0.5 to 3.2.2

Could you verify again with the new 3.2.1 please ?

comment:23 Changed 21 months ago by bruno

  • Resolution set to worksforme
  • Status changed from assigned to closed
Note: See TracTickets for help on using tickets.