Private Cloud Appliance
 
Private Cloud Appliance X5-2 Hardware
 

PLA-Support>Sun Systems>x86>Engineered Systems HW>SN-x86: OVCA
 

In this Document

Goal

Solution

Oracle Confidential PARTNER - Available to partners (SUN)
Reason: Internal Process

Applies to:

Goal

For certain few customers, this procedure is used to change management node Raid 0 configuration to Raid 1 configuration.

Solution

For certain few customers, this procedure is used to change management node Raid 0 configuration to Raid 1 configuration.

This process takes about 1.5 hours to 2 hrs per node.

First step is to login to ovcamn05 ILOM.
If you are connected to port 19 of the ES1-24 switch that you used to installation ssh as root to 192.168.4.103.  

Open a remote console  > start /SP/console

At login prompt, log in as root, with Welcome1 password.

Double check to make sure that the raid configuration is a Raid 0 as expected:

At the linux prompt type #  /opt/MegaRAID/MegaCli/MegaCli64 -ShowSummary  -a0

You should see the output showing the disks and the last few lines will show:

Storage
       Virtual Drives
                Virtual drive      : Target Id 0 ,VD name
                Size               : 2.180 TB
                State              : Optimal
                RAID Level         : 0                <<<<<<<<<<<<<

If it says RAID Level 1, then this node is fine and needs no change.
Be sure to do the above for management node ovcamn06, both should be the same.

Hint: if you open another ILOM session you will be able to view the console in one, while issuing your start and stop sys commands in the other.

At the ILOM prompt do a  > stop /SYS  
Give the node a few minutes to come down.

Next do a > start /SYS
Watch the console and when prompted enter the LSI configuration utility by pressing Ctrl+R  

This will bring up LSI gui shown below.
On the right side, under "Virtual Drive 0" it shows the config as a Raid 0 

Use the tab key to navigate across the top menu. Use the arrow keys to move the highlight to the item to be selected.
As shown below, under VD Mgmt tab, select Clear Configuration.
To get the popup screen shown below, you need to press F2 when the blue highlight is on the first line "LSI MegaRAID 936i-8i".
If the F2 does not work due to your terminal emulation, try pressing the Esc key quickly followed by the 2 key.  

The next step is to create a new virtual drive. Note that the two drives will show Unconfigured
Select Create Virtual Drive

Next select the Raid level from the drop down, select Raid 1
You will need to press the return key when the highlight is on "RAID-0" to see the drop down menu shown here.

The screen should now display as follows.  

You will need to select both drives by using the space bar to select them which will put an 'X” next to each one. 
Then select OK

The screen below should show the configuration as seen below.

Use the tab key to move to Ctrl Mgmt at the top bar, we need to make the device bootable.
It is not shown here, but the boot device may indicate 'None'.
In any case, using the pull down menu, select the VD you created under Boot device as shown here: 

Navigate to the Apply button and press return:

Next press escape to exit out:

Upon exit it will tell you to do a Ctrl-Alt-Delete to reboot, 
You may not be able to send Ctrl-Alt-Delete from your terminal emulation.
If not, you can do an > reset /SYS or simply push the button on the front of the management node.  
Watch the console and you should see the new configuration appear as below

Note:
The ESC may not allow you to exit to the Ctrl-Alt-Delete screen.
In this case, go ahead and restart the system via CLI (a new putty ILOM login to stop then start /system), or just use the power button.
The LSI configuration should be saved as you move through the utility.

At this point it will take about an hour to 1½ hrs to complete, as it will reboot several (about 3) times.
Before the final reboot, you will see the following text:

Unmounting update install
Unmounting update install
Installation process complete.
Installation process complete.
System is customer ready at next reboot. <<<<<<<<
System is customer ready at next reboot.
Rebooting now to finish update.
Rebooting now to finish update.

It will reboot again and you will see screen below shows the output at the end.  
Look for the line “Joining domain ovca (0 1) 2 nodes.”
The last line you might see is the bond2 link becomes ready, then we had to hit return to get the login prompt. 

This mgmt. node will now be the not-active node as ovcmn06 would have taken the master role when we did the first stop /Sys if it wasn't already the master.

You can check it by running # ovca-check-master
NODE: 192.168.4.3  MASTER: False

[Note this is the expected output for the standby node.]

# service ovca status
Checking Xsigo Fabric Manager tomcat: Stopped
MySQL is not running                                       [FAILED]
Oracle VM Manager is not running...
tinyproxy is stopped
dhcpd is stopped
log server (pid 5449) is running...            
remaster server (pid 5452) is running...        
http server is stopped
taskmonitor server is stopped
xmlrpc server is stopped
nodestate server is stopped
sync server is stopped

You are done with management node ovcamn05 at this time.

Now you can use this same procedure to change the Raid in ovcamn06 which will be the same other than the IPs  for the host and ilom connections.

This solution has no attachment