Replacing bad disk on R620 with different size disk?

Hi, could you provide your RAID controller information please.

Good to know all that! I wonder if my "slower" drive won't work at all or it will just slow down the entire filesystem to its speed ( 7.2k vs 10k ) ?

Thisisalloneword,

To answer your question, no you would not be able to use that drive to replace the Predicted Failure drive. The reason being is that while you don't have to have an exact match for the drive, you do have to use a drive that is either the same or that is larger and faster than the original, you can not use a drive that is slower or smaller.

Now on a side note, based on your pictures, it looks like the Predicted Failure drive you referred to is still in an ONLINE state. In order to replace an Online Predicted Failure, you will need to Force that drive Offline before removing it for replacement, this will keep the controller from transferring the predicted failure to the new replacement drive. 

Let me know if this helps clarify everything.

Hello thisisalloneword,

When we have a predictive fail drive that is still an online member of the array, there are 2 things that should be done before pulling and replacing: using OpenManage Server Administrator (OMSA)  or  the Controller BIOS

1.Consistency check on the virtual disk

2.When completed consistency check then Put drive offline before replacing

Then replace drive and rebuild should start automatically and if it does not you can set it as  hot spare and the rebuild should begin.

Ok, interesting to learn. How serious a problem is this given that the disk is still "online" . Do I need to to wait till 3 of the 4 disks report errors and go offline before I replace some? The file server is a bit obsolete and may get decommissioned in another year.. 

The controller won't allow it, it will show ready, but won't be listed when you try to rebuild the Virtual Disk as it isn't compatible.

Now you are close though, As when you put a larger or faster drive in, the controller will rebuild it as the same as the drives used to configure the Virtual Disk, but it won't slow the whole Virtual Disk down to match a replacement drive.

Thanks. I have ordered the identical replacement drive but am not sure how to access the "OpenManage Server Administrator" ? I do not see it anywhere in the iDRAC webpage for Storage.

Thisisalloneword,

Openmanage Server Administrator is its own systems management software that you would install in the OS, so it wouldn't be contained in the iDrac.  If installed it is normally an icon on the desktop called Server Administrator.

If you have the icon then you can us OpenManage and then go to Storage then Physical Disks, in the drop down of the drive in question you would select Force Offline. Once offline then remove it and wait 20 seconds and insert the replacement.

If you don't want to go that route, and you can schedule downtime, then you can power off the server and then remove the drive, afterwards boot back to the OS and then insert the replacement. Do NOT add/install the replacement drive to the server until you boot it back to the OS. 

Once in the OS, If the replacement drive doesn't start rebuilding when you insert it, then assign it as a hotspare and that should start the rebuild.

Let me know if this helps.

You run the risk of the bad blocks transferring to the other drives in the Virtual DIsk, which can lead to a double fault or a punctured stripe, possibly corrupting the data. Now just removing that single drive still leaves you at risk, if another drive was to fail it would fail the Virtual Disk. 

So there are risks present, but it is your call.

Many thanks for these helpful tips. Seems there are some risks/challenges with this replacement. Before I commit to doing it, can you advice on how critical this problem is? Right now only one disk is reporting errors. Will the drive still work if two or three out of the four start reporting  errors? We may retire this hardware next year. 

I finally got around to replacing the bad drive today. As I suspected, it was not going to be straightforward. I powered down the server and removed the faulty drive. I then booted back to the OS but got an error about one virtual disk being "Degraded".  It then would not boot the OS properly. The OS said there was no controller and would only let me login to some "emergency mode"..

I then attempted to put the faulty drive back in and boot the OS. THis time it complained about some "foreign configuration found on adaptor" and again would not boot the OS properly ( only emergency mode )..

At this point I am ok with just formatting all four drives, and reinstalling the OS. I don't need any data off the disks as I already did backups.

See screenshots below:

I gave it some more thought and am not ready to go the "nuclear" route of reinstalling the OS. Would like to try and use the data on the drives but not sure how to fix this "degraded" virtual drive. It is Raid 10 so should still be ok if one drive is missing. Here's what I see in the BIOS config and iDrac:

Thisisalloneword,

I would recommend at this point to hold off until we get a better idea as to what the controller and virtual disks are doing. Would you follow the link here and go through thr walkthrough on pulling a tty log from the controller, after you pull the log would you upload it to upload. dell. com and private message me the svc tag you used to upload it? That way I can get a better idea as to what is causing the issue. 

Let me know.

You would need to use a USB, as seen above step 1. 

Let me know if it proceeds.