Notifications
Clear all
Investigating Problems with Disks
10
Posts
2
Users
0
Reactions
6,632
Views
Topic starter
I've had a drive with an error and I'm going to replace it.
I need to make sure I understand the process correctly.
I have SoftRaid running on a MacPro (cheesegrater) as internal hard drives (1+0) and a PCIE SSD card for booting.
Process:
Identify which drive is failing (in the case it's saying bay2)
Shut down the MacPro
Open the case and replace the drive with the new one.
Boot back up.
Do I need to do anything else or will it rebuild as normal.
It's a 2TB RAID about half full, so I assume the rebuild will take a while.
How will I know that its done?
Posted : 21/05/2019 10:41 am
There are a few additional steps:
We also have a video to help walk through this:
https://srforums.wpengine.com/pages/videos.html#setup
Go to the support video on replacing a drive.
I will assume the drive is reallocating sectors or failing SMART, not just an "IO Error", which is not necessarily a failed disk and should be investigated further.
You need to insert the new disk
Certify disk is highly recommended (will take several days)
Initialize the disk with SoftRAID
"Add disk" and then you are done.
You should optimise for workstation or server for faster rebuild times.
Posted : 21/05/2019 12:46 pm
Topic starter
Thanks for the reply and the video.
As I have a MacPro cheesegrater I need to shut down the Mac where it says 'power down the RAID'?
I looked at the log and it said the drive needs replacing.
In terms of certifying - will the RAID be available for normal working whilst it's doing this?
Posted : 22/05/2019 3:28 am
Does the drive have a status of failed SMART, or reallocated sectors? Or is it "just" an IO error?
(trying to be sure you don't have a "false alarm", that the drive is definitely in the failing stage.)
Do you know which bay it is in? SoftRAID helps by numbering the drive bays, but you need to make sure you have the correct drive in. If you start up with only 2 disks in, then you can start to identify the disks more easily.
Yes you can access the volume while you certify. You won't be redundant, but your volume will be available while you certify.
Posted : 22/05/2019 12:20 pm
Topic starter
Does the drive have a status of failed SMART, or reallocated sectors? Or is it "just" an IO error?
(trying to be sure you don't have a "false alarm", that the drive is definitely in the failing stage.)
I haven't certified these disks (not sure if that makes a difference), it doesn't say anything next to the smart status or reallocated sectors, but in the console log it says it needs replacing.
Do you know which bay it is in? SoftRAID helps by numbering the drive bays, but you need to make sure you have the correct drive in. If you start up with only 2 disks in, then you can start to identify the disks more easily.
It says (bay2) so I'm assuming that is correct and it's in bay2?
Posted : 23/05/2019 1:25 am
Topic starter
So I ran a validate on the volume before doing anything else and it said that it failed variation and should be replaced.
I assume that's not a false alarm.
Posted : 23/05/2019 2:57 am
Correct.
Posted : 23/05/2019 12:10 pm
I think the bay numbers start from the back, but am not certain.
Posted : 23/05/2019 1:25 pm
Topic starter
Thanks for the update - just as a closure to this thread.
The bay numbers do match up as what is shown in SoftRaid.
Following the process was really easy, shutting down the MacPro, pulling the drive, booting up and rebuilding went without a hitch.
Posted : 24/05/2019 3:01 am
Fantastic!
Posted : 24/05/2019 1:02 pm
