[ASK] The safe way for fallback if update in CCM fail

Answered Question
Jan 11th, 2009

Hi, i want to ask if it is the safe way if i do fallback in MCS (HP Blade) using RAID feature?

I was told by server guy that fallback can be done using mirroring by RAID 0+1.

The sequence is like below :

Before updating CCM :

1. shut down MCS

2. Remove one of two harddisk from its tray as backup

3. Turn on MCS

After updating MCS, if the update succeed :

1. just put the backup harddisk back to its original tray to get mirrored from updated harddisk.

After updating MCS, if the update fail :

1. Shut down MCS

2. Put the backup harddisk back to its original tray.

3. remove the updated harddisk (fail one) from its tray

4. turn on MCS

5. Put the failed harddisk back again to its original tray to get roll back

Is the sequence action already right?


I have this problem too.
0 votes
Correct Answer by Adam Thompson about 7 years 9 months ago

Yes, your sequence of events is correct for both successful and failed updates. I have had to go through both scenarios and have had no issue with any of our servers.

Please rate if helpful

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
Jaime Valencia Tue, 01/13/2009 - 04:58

if you're using a linux appliance there is no need as there is a switch version option



if this helps, please rate

Charles_Chi4 Tue, 01/13/2009 - 07:20

sorry not telling u that the object is MCS 7825-H1-1PC with CCM 4.1.(3) installed in it.

And i just saw that the harddisk LED is not working and there's a note stick on it says that the SATA hot plug and LED indicator is not supported by the controller.

Correct Answer
Adam Thompson Mon, 01/19/2009 - 10:09

Yes, your sequence of events is correct for both successful and failed updates. I have had to go through both scenarios and have had no issue with any of our servers.

Please rate if helpful

bas.morgenrood@... Wed, 01/21/2009 - 01:54

Hi Charles,

Although your squence is OK, I still have to make some remarks. I spend a lot of time in solving a RAID issue, after the motherboard has been changed by HP due to a defective on-board RAID controller.

Setup was the same as yours, MCS7825-H1 (SATA drives) and CM 4.1.3

Both Cisco and HP could not give good working instructions how to restore RAID without doing a re-install. They told me to perform a clean install. RAID config and options are not documented for a MCS7825-H1


I found out (althought Cisco mentioned it was not possible) that you can swap a drive from left to right and opposite. But you have to pay attention to the warnings messages during boot and accept the new situation.

Also you can rebuild from right to left, but you have to configure that from the build-in RAID BIOS during the boot stage.

At the end I rebuilded the system from scratch for my customer, but that would have not been necessary.

It is expected that the LED's on the HD,s are not working, it is not supported on that type of server.

Further I found out that the server always tries to boot from the left disk, so if you have a broken (just data not Hardware) disk in the left slot (as I my case), and a good one in the right hand slot, the system won't boot. (In my case it entered a blue screen)

It is however possible to overcome that situation, but you have to pay attention. You can delete the array config with the BIOS utility of the on board RAID controller. (press CTRL+A) during boot when you see the message on the screen). Then delete the array, you will get a big warning that data will be destroyed, but it is only the RAID configuration what will be deleted.

The next step is very important, it will ask you if want to delete the partitions as well, CHOOSE NONE, otherwise you will destroy all data on the disk and than you will have todo a clean install.

Then make a new array, name it for example Array_1, and choose the good disk, (for example the right hand one) as source disk. If you accept it, the rebuild will start. It takes about 45-60 minutes to rebuild.

If you want to pull a disk before the upgrade, I would pick the right hand one, then boot from the left disk and do the upgrade. If the upgrade succeeds, switch the system off and install the right hand disk.

Boot the system, and open after the system has been booted, HP Storage Manager from the Start menu. (Pay attention, OS 2000.4.1 does not install HP storage manager, you have to upgrade to 2000.4.5 or higher)

Double click the RAID controller icon. You will see that disk in slot1 (right hand slot) is degraded or failed. Then choose rebuild option and follow the on screen messages.

If this is not working for you'll have to delete the array as described before.

Now in case yo have to switch back to the non upgraded version on the right hand disk. I advise to delete the array from RAID BIOS with CTRL+A, don't delete the partitions, make a new array and rebuild from right to left. (right hand disk as source disk)

slot0=left, slot1=right




This Discussion