PowerEdge R620 - drive predicted failure

Hello

I responded to another post about predictive failure earlier. Most of what I typed in that reply would apply here. There was also an earlier post where I talked about internal speed and transfer rates in relation to virtual disk requirements.

https://www.dell.com/community/PowerEdge-Hardware-General/acceleration-of-the-rotation/m-p/6111846

https://www.dell.com/community/PowerEdge-HDD-SCSI-RAID/disk-speed-and-capacity-question/m-p/6111131

You can find details about replacement procedures and supported configurations in the controller manual.

http://www.dell.com/storagecontrollermanuals/

The virtual disk rules are mostly industry standard. RAID 5 will use the space of the smallest disk in the array, so any replacement drives need to be equal to or greater than the smallest disk. I'm not aware of any of our controllers that support drive mixing within a virtual disk, so the replacement must be of the same type(SAS, SATA, SSD).

Thanks

Thank you for this great information.  I have ordered the exact same drive that was in the system  . 

Something I am still not clear about (since this part of the system is RAID-1)....is how do i replace the bad drive?  I saw something that said to take the bad drive offline....is that from the server manager?  Do I then remove the bad drive and put in the new drive and it should start the rebuild?  Or do I need to power the system down.  Replace the drive and restart.

Thanks again for all your help.

If the drives are hot swap then it is best to replace the drive while the system is on. You will need to use one of our storage management applications like OMSA to offline the drive.

How do I know if they are hot swap?  

It will usually start the rebuild automatically, but if it doesn't, you will need to add it as a hot-spare. If the drive shows up as Foreign, you will need to CLEAR that foreign configuration first, then assign it as a hot-spare. (NEVER set a newly-introduced drive "online".) Server Manager has no visibility into hardware RAID drives, so you will need to use the OpenManage Server Administrator application.

It's ok to do while the server is in use, but there will be a performance hit while it rebuilds, so if it is heavily used and is critical, you may want to wait until after hours.

OK yes they are accessible from the front.  So I can go into the Server Manager take the failed drive offline.  Remove it then insert the new drive.  Will automatically start the rebuild?  do I need go in and set it online. 

Is it OK to do this while the server is being accessed?  Or should I wait until after hours?

I apologize for so many questions but I am such a novice when it comes to this I don't want to mess up these systems.

I am using the OpenManage Server Administrator.  I was saying the wrong thing.  I have ordered the drive and have not received it yet.  It is supposed to be a brand new drive.  Will I clear the foreign config from OSMA (if it comes up with that)

If the drives are accessible from the front of the machine (in trays/caddies), then then are hot-swappable.

Sorry ... because Windows Server's management interface is called Server Manager, when people say that, I assume that is what they meant :)

If it is new, it will probably rebuild automatically, but if it is refurbished or bulk (basically new but cannot be called "new"), then it may or may not. If it doesn't and shows up as Ready, then assign it as a hot-spare. If it doesn't and shows up as Foreign, then you must CLEAR (do not Import) that foreign configuration in the dropdown of Available Tasks in the controller section of OMSA, THEN you can assign it as a hot-spare.

OK thanks again for all your information.  The new drive should be here today.  So hopefully it goes off without a hitch.

Steps

• Take bad drive offline
• Remove and insert new drive
• Rebuild should start
• If not, in OMSA look to see if it comes up as a Foreign Configuration
• If it does change it to Hot Spare
• Then rebuild?

Again I apologize for all the question but this is not my area of expertise.  Just need a little reassurance.

---

@braunhtf wrote:

OK thanks again for all your information.  The new drive should be here today.  So hopefully it goes off without a hitch.

Steps

• Take bad drive offline
• Remove and insert new drive
• Rebuild should start
• If not, in OMSA look to see if it comes up as a Foreign Configuration
• If it does change it to Hot Spare
• Then rebuild?

Again I apologize for all the question but this is not my area of expertise.  Just need a little reassurance.

---

Those steps look fine, but I would add two more at the very beginning.

1. Make sure you have a valid tested backup before doing anything
2. Run a consistency check before replacing any online drives, if your controller and RAID level support it.

Before making any hardware or software changes you should get a valid tested backup. Running a consistency check will allow the controller to make. If it comes up as a foreign configuration then you will need to either import or clear. In your situation you most likely be clearing. It is unlikely the replacement disk will come up as foreign unless it is being moved over from another system. If you encounter that situation you may want to just stop and ask for help if you don't know.

A quick update - I replaced the drive and it started to rebuild and has completed successfully.  Thanks again for everyone's input and helping me through this issue.