P420i RAID10 failed HDD, doesn't rebuild after HDD replace

[dwma]

Hello,

We have problem with Smart Array P420i on our ML350p Gen 8 server. One drive in RAID10 has failed (4x WD Purple 4TB), so we've replaced it with the same drive.

The problem is, that after HDD replace it still shows up as failed (Box 1: Bay6 is bad or missing), so we've tried another WD Purple 4TB (NEW!) drive, and the same.

HDD in Bay6 is visible as "Unassigned drive", I tried to configure it as a spare, but nothing works. I've also tried restartin server several times, but I've not received any info like "Press F1 to start rebuild".

I've upgraded P420i firmware from 2.14 to 8.00, but it still does nothing.

[Fritsch James]

Hello,

ypu should be using original HP drives with HP Firmware. The Smart Array controllers are not optimized for the use of foreign drives - especially the "WD Purple" are not recommended for RAID use, you should use "WD Red Pro" or "WD Gold" drives instead, but original HP labeled drives would be better at all.

[dwma]

This is not an answer. Also HP is not a disk manufacturer, because mostly it's branded Seagates etc.

We use this disks for Survalliance server, and the replacement disk is detected by P420i (because I see it in slot6, but at "unassigned drives").

So my question is how to force Smart Array to use this "unassigned disk" to rebuild the RAID10?

[Fritsch James]

Yes you are right, the HP disks are manufactured by Seagate or Western Digital - but they have their own special firmware which is built by HP and therefore behave different than the WD or Seagate branded drives on the retail market.

It is also true that the WD Purple series drives are not suitable for use with a hardware raid controller.

Which operating system do you use for your "Surveillance Server" ?

Do you have HP SPP components installed?

[dwma]

I came here to find a solution to our problem, not to hear reply, that seems to be some "commercial vaporware" (as someone perfectly described it somewhere on community).

Disk is disk, if it's "not suitable" why it's working perfectly from 2 years on it? Next thing, do you have some document saying, that WD Purples are not suitable for HW Raid? If yes, please show it to me.

So please just stop it, I need working solutions, not just some "vaporware thesis". END OF TOPIC.

-----------

Our ML350p G8 is running ESXi 6.0 and Windows Server 2008 R2 on top of it.

I've updated all server components FW via the latest SPP for this server.

[Jimmy Vance]

A couple of things can be at play here.  If the system isn't seeing a replacment drive then there could be a backplane issue. If it is seeing the replcement disk as a new device then there is a good chance that the capacity isn't exactly the same.  Some drives even though they are labeled as xTB or xGB may not have exactly the same capacity as another drive marked the same.  For drives used as RAID drives, the capacity of each device must present the same capacity (or larger) than the smallest device in the array

[dwma · Accepted solution]

I've deleted old RAID10 array, and recreated a new one (with newly added disks) all working fine, so there were not backplane issue, and the replacement HDDs were the same size (im 100% sure about that).

Dunno why this happened, and why replacement disk were not detected properly/RAID were not rebuilding.

[Jimmy Vance]

---

@dwmawrote:

I've deleted old RAID10 array, and recreated a new one (with newly added disks) all working fine, so there were not backplane issue, and the replacement HDDs were the same size (im 100% sure about that).

Dunno why this happened, and why replacement disk were not detected properly/RAID were not rebuilding.

---

Glad you got it sorted out. Was probably a difference in size. Even though a drive may be labaled as X, if the actualy capacity is just a bit lower than drives already in an array, it won't be added.  If one disk is just a bit lower when you create a new array, all the drives will be sized down to meet the actual capacity of the smallest drive