Many HPE ProLiant Gen11 servers ship with HPE MR Gen11 controllers instead of, or next to, the SR (Smart Array) family. MR controllers are tri-mode (NVMe, SAS and SATA), run in mixed mode (RAID and JBOD at the same time), and use their own tools. This guide follows the HPE MR Gen11 Controller User Guide and the MR Storage Administrator (MRSA) User Guide to create logical drives, add hot spares and set cache policies.
Before you start
- Identify the controller: System Utilities shows it as HPE MRxxx Gen11 under System Configuration, and iLO lists it under System Information > Storage.
- SSA and ssacli do not work with MR controllers. HPE's guide states that MR controllers are managed with MR Storage Administrator, StorCLI or UEFI System Utilities.
- During initial provisioning, HPE says to configure the controller in UEFI System Utilities. MRSA and StorCLI are available in the Service Pack for ProLiant (SPP).
- Back up any data on drives you plan to use. Creating a logical drive permanently deletes data on its drives.
How MR differs from SR (Smart Array)
| Topic | MR Gen11 | SR (Smart Array) |
|---|---|---|
| Tools | UEFI System Utilities, MR Storage Administrator, StorCLI | UEFI System Utilities, SSA, ssacli |
| Free drive states | Unconfigured Good (hidden from the OS) or JBOD (exposed to the OS) | Unassigned |
| Spares | Global or dedicated hot spares | Dedicated or auto-replace spares |
| Write cache | Write Through, Write Back or Always Write Back per logical drive | Controller cache ratio and per-drive accelerator |
| Portability | A RAID configuration created on an MR controller cannot be moved to an SR controller | — |
The MR 200 series (MR216i-o and MR216i-p) has no cache and supports RAID 0, 1 and 10. The 400 series (MR408i and MR416i, -o and -p) has flash-backed write cache and supports RAID 0, 1, 10, 5, 6, 50 and 60. The optional energy pack is not supported on MR 200 series controllers.
Step 1 — Check drive states and Auto-Config
In mixed mode, each drive is a logical drive member, Unconfigured Good, a spare, or JBOD. Only Unconfigured Good drives can be used in a logical drive. When a new drive is added, the Auto-Config setting decides its state: JBOD or None (stays Unconfigured Good). HPE documents the defaults as JBOD for MR 200 controllers and None for MR 400 controllers; Auto-Config can be changed from firmware 52.26.3-5250.
To convert JBOD drives to Unconfigured Good:
- Restart the server and press F9 at POST to open System Utilities.
- Select System Configuration > (MR controller) > Main Menu > Drive Management to see each drive's state.
- Select Main Menu > Configuration Management > Make Unconfigured Good.
- Select the drives, click OK, then Confirm and Yes.
Warning: A JBOD drive is visible to the OS. Converting it to Unconfigured Good removes it from the OS, and creating a logical drive on it destroys its data. Check that no OS volume or datastore uses the drive, and back up first.
Step 2 — Create a logical drive in UEFI System Utilities
- Open System Configuration > HPE MRxxx Gen11 > Main Menu > Configuration Management > Create Logical Drive.
- If JBOD drives are listed, decide whether to convert them or click Skip.
- Choose the RAID level and set Select Drives From to Unconfigured Capacity (new drives) or Free Capacity (unused space on an existing array).
- Open Select Drives: filter by media type (SSD, HDD or Both), interface (SAS, SATA, NVMe or All) and logical sector size (512 B, 4 KB or Both), select the drives and click Apply Changes, then OK.
- Set Logical Drive Name, Size, Strip size, Read cache policy, I/O Policy, Access policy, Drive cache and Default initialization (No, Fast or Full).
- Click Save Configuration, review any warnings, then click Confirm and Yes.
Fast initialization overwrites the start and end of the new logical drive and finishes in the background; Full initialization overwrites every block, and you cannot write data until it ends. Both destroy existing data on the logical drive.
Step 3 — Or use MR Storage Administrator
MRSA is a web application. Install it in the OS from the SPP; by default it listens on web server port 2463, so browse to the server IP address followed by :2463. It is also available offline: press F10 and select MR Storage Administrator, or choose it under Perform Maintenance in Intelligent Provisioning.
- Simple Configuration (Configure > Simple Configuration): choose RAID 0, 1, 5 or 6, the number of volumes and their capacity, and optionally tick Assign Spare. The system picks the drives.
- Advanced Configuration (Configure > Advanced Configuration): choose the RAID level, click Add Drives and optionally Add Spares, then Add Volumes to set name (up to 15 characters), strip size (64 KB to 1 MB), initialization, read policy, write policy, I/O policy and drive write cache. Click Finish. The guide lists RAID 10, 50 and 60 (spanned arrays) for this option.
Step 4 — Assign hot spares
A global spare replaces a failed drive in any array if it is the same drive type and at least as large; for RAID 0 it activates when a member reports a predictive failure. A dedicated spare protects the arrays you choose. See global vs dedicated spares.
- UEFI: Main Menu > Drive Management, select the drive, choose Assign Global Spare Drive from the Operation menu, click Go and confirm.
- MRSA: on the Controller dashboard open the Drives tab, expand Unconfigured Drives, select the drive and choose Actions > More Actions > Assign Global Spare Drive or Assign Dedicated Spare Drive (then pick the array).
Under Controller Management > Advanced Controller Properties > Spare you can enable Persistent Spare (a drive that replaces a spare in the same slot is automatically configured as a spare), Replace Drive (copy data from a spare back to a drive) and Replace Drive on SMART Error.
Step 5 — Cache policies and the energy pack
- Write Through does not use the controller cache. Write Back uses it and switches between write back and write through depending on controller status. Always Write Back uses it regardless; if the energy pack is absent, the firmware forces Write Back.
- Read ahead or No read ahead, I/O policy Direct or Cached, and Drive cache Unchanged, Enabled or Disabled are also set per logical drive.
- MRSA notes that when the HPE Smart Storage Energy Pack is not optimal, the WriteCache policy is disabled. Check it in iLO under Power & Thermal > Power.
Monitoring and rebuilds
When a drive in RAID 1, 5, 6, 10, 50 or 60 fails, the firmware rebuilds onto a spare or replacement drive automatically. Watch progress in MRSA's Background Processes in Progress window. The rebuild rate (1–100) is under More Actions > Set Adjustable Task Rate in MRSA, or Advanced Controller Properties > Task Rates in UEFI; a higher rate can slow system I/O. For command-line checks, see HPE StorCLI basics.
Warning: Clear Configuration deletes all logical drives and spare drives on the controller. Deleting or initializing a logical drive destroys its data. Back up first and confirm the controller and logical drive before you click Confirm.
Frequently asked questions
Can I manage an MR controller with SSA?
No. HPE states that MR controllers cannot be managed with Smart Storage Administrator; use MRSA, StorCLI or UEFI System Utilities.
Can I move MR arrays to a Smart Array controller?
No. The MR Gen11 guide notes that a RAID configuration created for an MR controller cannot be moved to SR controllers. Plan a backup and restore instead.
Checklist
- Controller model and series (200 or 400) identified.
- Target drives are Unconfigured Good and hold no needed data.
- Logical drive created with the right RAID level, strip size and initialization.
- Global or dedicated hot spare assigned.
- Write policy and energy pack status checked.
- Rebuild rate reviewed and MRSA alerts configured.
Comments
0 total · 0 threads