RAID 5 Array 4 drives 1 failure - 3 working

T

The Abyss

Active Member
Joined
Jul 16, 2014
Messages
94
Reaction score
1
Location
EveryWhere
Hello Every one

I have questions about RAID 5 and rebuilding.

Server has been running now almost 3 or more weeks in a degraded state, 1 failed drive removed.
Now to insert a new same capacity SAS drive to the array.

I have the server raid M5014 sas/sata controller. Failed drive has been removed, now to re insert new drive and boot into raid controller

First time adding drive in windows, intel raid web console detected drive added it to raid, told it to go online and the server froze... Rebooted and removed drive. Eventually booted up again.

Does it matter if you rebuild a RAID 5 array in the RAID bios without updating the firmware, as i cannot find updated firmware for said controller.

I believe adding the drive and rebuilding this way is ok.

Please can any one feel free to let me know their thoughts and experiences.

Thank You in advance.
 
BEFORE YOU ADD A DRIVE - do a FULL backup BEFORE inserting the new drive.

The reason for this is why - if another drive fails during a rebuild, and you're on a RAID5 array, then you're stuck.

If on RAID6, it will still continue rebuilding (RAID6 can handle two failed drives), but if a second drive also fails, then you're stuck.

So my recommendation is to do a full backup while the RAID is still working even though one (or two for RAID6) drives have failed...

Should the RAID fail during rebuild, you will still have your data, and won't need to pay top $$$ for a RAID recovery specialist - these are 50/50 at recovery.

I haven't updated my RAID array firmware (also using Intel) and never had any problems, even when hotswapping the faulty drive with a new drive (AFTER doing a full backup).

FYI if you want to do hotswapping of hard drives, you'll need to have a proper hotswap RAID cage, or else you have to cold swap the drives.


I cannot emphasize the backup enough, heard too many horror stories of a RAID5 failing during rebuild... and it do happen.
 
Thank you Librarian.

I have got daily incremental backups from VEEAM, now doing a full active backup of both my VM's

Is it blind to have faith with 3 working drives the 4th should rebuild ok?
 
What server platform is this? The card model you've mentioned seems to return as an IBM branded card, but you're using Intel web console?

Is the replacement drive the exact same model as the existing disks? If not what did you have and what are you replacing it with?

Most of these types of cards have a console you can access at boot time. See if you can initialise the rebuild from there.

RAID 5 with 1 failed disk? You're just one more failure away from total data loss. Hope your backups are working and tested.

Your brute force option of course if you are running a Hypervisor and have Veeam backups is to back it all up and create the RAID array from scratch.
 
onexios said:
Thank you Librarian.

I have got daily incremental backups from VEEAM, now doing a full active backup of both my VM's

Is it blind to have faith with 3 working drives the 4th should rebuild ok?
Click to expand...

Historical data has shown that there is a fair chance (for all you keyboard warriors out there, I said fair and not guaranteed) of a second drive failure occurring within weeks of the first drive failure, for large capacity arrays. Your 4th drive would probably fail to rebuild due to another drive failure that kills the entire array.
 
Last edited:
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X