MXe-III RAID Drive replacement

[AH64Armament]

I have a controller that has a failed drive.
The tech handbook method for replacing a single drive failed.... or I missed a step.

Drive 1 failure. Drive 2 operational

Removed Drive 1 from the system. Replaced the drive in the carrier and reinserted the carrier into the system. According to the THB, the RAID should have started to rebuild.

Instead, from the show status raid and show status redundant, the Drive 1 indicates ID Mismatch - shows the ID of the drive that is in there, and the ID of what is expected.

I have already followed the process to clear the sockets - shut the system off. Remove both drives. Power back on, monitor the system for "Sockets Cleared". Power back off. Insert the good drive. Power the system back on and let it come all the way up. Then insert the new drive. Raid rebuilt itself.



Here's the notes from the THB... I am thinking there is a missing step between 4 and 5.
MXE III (TWO HARD DRIVES)
If the MXe III has two hard drives in a RAID (Redundant Array of
Independent Disks) configuration, you can replace one or both of them.
• "Replace One Hard Drive in an MXe III" on page 244
• "Replace Both Hard Drives in an MXe III" on page 245
REPLACE ONE HARD DRIVE IN AN MXE III
Replace one hard drive in a RAID configuration if only one hard drive is
defective and the system is no longer under warranty.
To replace one hard drive in an MXe III:
1. Release the retaining screw securing the hard drive carrier to the
controller.
2. Slide the defective hard drive out of the hard drive carrier.
3. Remove the four screws securing the hard drive to the hard drive carrier.
4. Slide the (replacement) hard drive or SSD into the drive carrier.
Note: If the system is under warranty, you must replace both hard drives.
Before proceeding with a warranty claim, contact Mitel Repair Services
(page 10) to obtain a Return of Merchandise Authorization (RMA) number.
Note: The defective hard drive is indicated by a flashing green Host LED.
Tip: Refer to Table 75 on page 377 for a complete description of LED activity.
245
Install and Replace Units
- Initially, loosely install the top two screws (Figure 72 on page 243).
Ensure the hard drive is correctly oriented (i.e., right side up as per
Figure 73 on page 243).
- Install and tighten both side mount screws (Figure 74 on page
243).
- Tighten both drive top mount screws, taking care not to twist or
bend the mounting frame.
5. Slide the hard drive carrier into the controller.
6. Push to seat the hard drive carrier into the hard drive backplane.
7. Tighten the thumb screw.
8. The rebuild process starts automatically. Mirroring is indicated by the
HD LEDs. The source drive LED flashes quickly (indicating that the
disk is being accessed) while the destination drive flashes slowly. The
rebuild is complete when the destination drive LED no longer flashes
slowly (approximately 80G per hour).

 

[AH64Armament]

It worked on the 2nd drive... so confusing.

The first drive - flashing SOS when I arrived. I pulled the drive and replaced it. I inserted the new drive to the system - flashed SOS and showed mismatch ID. I went through and powered the system off... removed both drives, cleared the sockets, power off, insert good drive (slot 2) and powered back on. Once up, I inserted the new drive into slot 1. The RAID rebuilt itself.

Once the RAID was clean, I removed Drive 2 and replaced it. I was expecting it to require me to clear the sockets again, but this time it started rebuilding the RAID.

This is one area that Mitel is very lacking on their documentation.

 

[nytalkin]

The THB is all wrong....Go to the KNOWLEDGE BASE and search for RAID MXeIII. This article is very clear on the process. If memory serves, it cover all controllers.

I suppose you're entitled to your opinion, I'm just not going to suppose very hard.

 

[AH64Armament]

I grabbed that article and it mostly parroted the THB. The info I was thinking as missing isn't there either.

But the good thing is that I when I replaced the old good drive with a new one, the RAID rebuild started.... I'm just confused as to why I needed to clear the socket when I replaced the old bad drive.

 

[jpruder]

I have yet to replace a failed drive without a reboot FWIE

 

[SXWizard]

The RAID controller stores the model number and a serial number of the disks in what is called sockets. This information is used to detect when a disk has been changed. When a RAID controller is shipped, this socket information is clear. There are times this information must be cleared, for example, when you want to install a hard disk that already has a different version of software or configuration, or if you start up the system with one drive installed and that drive is defective.
In a SATA system, you may need to clear the RAID controller sockets if one (only one hard drive installed) or both hard drive (two hard drives installed) LEDs flash the Morse code for SOS: a repeating pattern of three slow flashes and three fast flashes.