ProLiant Servers (ML,DL,SL)
cancel
Showing results for 
Search instead for 
Did you mean: 

Best Practice for Lab Disk Setup

 
SOLVED
Go to solution
Walt
Advisor

Best Practice for Lab Disk Setup

Hello. I have an HP Proliant ML370 G4 Server with Smart Array 641, configured with 3 identical Ultra-320 SCSI disks as RAID-5.

This is our office's primary server domain controller.

I've purchased an additional identical disk to use in the same system (hopefully) as a lab system on the weekends and I've run into my first concern.

My plan was to simply remove the 3 "Live" disks and put them aside, and boot with my new "Lab" disk, but now I've realized that I might need to overwrite the current Smart Array configuration in order to use the lab disk, as the the only disk.

I did not proceed.

Are my suspicions correct? If I were to delete the current logical drive, will those 3 SCSI RAID 5 disks become useless?

What's my best option to ensure these 3 disks remain as-is, so I can replace them at the end of the day, and dont lose any data?

Thanks in advance for any suggestions you may provide.


2 REPLIES
Brian_Murdoch
Honored Contributor
Solution

Re: Best Practice for Lab Disk Setup

Hi Walt,

The Raid configuration is held on the disks in the RIS tables (Raid Information Sector). This information is saved on the disks when you configure them using F8 during POST or by running the Array Configuration Utility (ACU).

So with this RIS config each disk in your current raid 5 knows which slot it was in and the serial numbers, size and slot numbers of the other disk members in the raid 5. This means that you can power down the server, remove these disks and a month later insert them back into the server and the controller will read the configuration back off the disks and know their setup. This is one of the things that makes the Smart Array Smart. It also allows controllers to be swapped without loss of configuration since it is held on the disks, not the controller.

If you power down and take out the three original drives then insert your single drive in bay 0 for example, it won't have any configuration data on it to start with (No RIS table information) so you will have to configure this single disk using F8 when prompted at POST as a RAID 0 (This is your only choice with one disk).
Once this is saved, the disk will be stamped with the RIS information and you can use it as you wish.

Once your tests are complete, remove it and put the original disks back in their original slots. On power up the controller will read the disk RIS tables and you will have your logical drive back again. Similarly the next time you insert the single disk back in the server on its own, the controller will know it's a raid 0 and find the logical drive ......simple.

Word of warning - Just in case you mix up the 3 disks when you lay them aside, mark their slot positions on them somewhere or use sticky labels. If you do mix them up the controller will actually sort it out for you (drive roaming is a feature of Smart Arrays) but it's best if you mark them first before removal.

I hope this helps,

Brian

Walt
Advisor

Re: Best Practice for Lab Disk Setup

Brian,
Thank you for the very detailed explanation. I appreciate your taking the time to do so.

This is great news. I was scared because I had a Dell server cause this problem for me in the past- what a nightmare that was.

Thanks again for the very good post.

~Walt