Need help: PANIC: Cannot find root disk controller after changing disks

Need help: PANIC: Cannot find root disk controller after changing disks

Post by Thomas Peppin » Tue, 13 Oct 1998 04:00:00

We have Openserver 5.0.0 on Compaq Proliant 5000 with two Smart 2 ida
controller, RAID1 array for system disk and RAID5 array for data.
After backing up all filesystems we enhanced system RAM from 256 to 512
MB and changed the disks of the RAID arrays. We also had to upgrade the
ROMpaqs for the system and the ida controllers.
Following the instructions in the SCO handbook, we made a fresh
installation with osr5.0.4, prepared the new disks and filesystems, did
not install any optional software like operating system services or
graphical environment.
In this state the system will boot without any problem.
We then restored the old root filesystem from tape and relinked the
kernel to get the old osr5.0.0 in place.
During reboot, the system panics since it cannot initialize the ida
The following messages occur:
Warning: ida0: IdaSend PCI ...Invalid command list address returned.
Request command list: 0x0014000000   Returned command list: 0xFFFFFFFFFF

Later on the following message is chown on the console:
Warning: hd: no root disk controller

We get the same error after restoring root- and boot-filesystem from

What do I have to do?? Is there any configuration file I have to change
manually before relinking the kernel ?


1. Duplicated boot disk panics; cannot mount root

My goal: duplicate a boot disk on a FreeBSD system.

My problem: after duplicating the disk, the duplicate will not boot;
after the generic kernel devices are configured, booting stops with a
panic, with the console indicating the root could not be mounted.

    npx0 on motherboard
    npx0: INT 16 interface
    panic: cannot mount root

    syncing disks... done
    Automatic reboot in 15 seconds - press a key on the console to abort

The hardware: Dell P75t, AHA2940AU, Seagate Fast SCSI-2 boot, Imprimis
SCSI-1 duplicate.

How the duplication was attempted (1):

    - duplicate disk came from a Macintosh
    - duplicate disk formatted with scsiformat
    - disklabel and boot written with disklabel
        disklabel -r -w sd1 auto
        disklabel -B sd1
        disklabel -e sd1
        - these steps sometimes result in console errors, including a
          lack of magic
    - newfs on /dev/rsd1[a,e,f,g]
    - swap drive ids and reboot

Since this failed, we also have How the duplication was attempted (2):

    - from MS-DOS, FDISK /MBR, and then made a FAT32 partition
    - from FreeBSD, change FAT32 partition to FreeBSD partition/slice
    - disklabel and newfs as above
    - swap drive ids and reboot to identical failure

Finally, How the duplication was attempted (3):

    - identical to process described in FAQ
    - swap drive ids and reboot to identical failure

There must be something obvious eluding me.  Please, what is it?  I'd
tear out my hair but I don't have much of it left.

I noticed in /dev I have no [r]sd1s[1234]* entries, though I have
[r]sd1* without slices.  Do I need to MAKEDEV these somehow or is this

No, I am not related to any Higher Source people.  :-/

2. email for multiple domains ?

3. Boot Disk Cannot Find Hard Disk.

4. Can I install freebsd without a host or domain name?

5. READONLY root file system

6. R: Boot Disk Cannot Find Hard Disk.

7. sendmail link to Msexchange

8. No root disk controller, No SCSI host adapter found

9. No root disk controller found (OS5)

10. OS5 ATAPI "No root disk controller found"

11. what different between boot disk, root disk and emergency disk???