A spare drive lets an HPE Smart Array (SR) controller start rebuilding a degraded logical drive without waiting for someone to swap the failed disk. SR controllers offer two spare types and two activation modes, and the choice changes what happens during and after a failure. This guide explains each option as HPE documents it and shows how to set it in Smart Storage Administrator (SSA), UEFI System Utilities and ssacli.
Before you start
- A spare must be an unassigned drive or a spare already used by another array.
- It must be the same type as the drives in the array (for example SAS or SATA).
- Its capacity must be equal to or larger than the smallest drive in the array.
- These steps apply to SR controllers. HPE MR controllers use global and dedicated hot spares instead; see RAID on HPE MR Gen11 controllers.
What a spare does and does not do
HPE's SSA CLI guide is clear on the limit: assigning spares lets you postpone replacing a faulty drive, but it does not raise the fault tolerance of a logical drive. A RAID 5 logical drive still loses data if two of its drives fail at the same time, however many spares the array has. A spare shortens the time the array runs degraded because the rebuild starts at once. For the general concept, see hot spares: global vs dedicated.
Spare types: dedicated or auto-replace
| Type | Behaviour (HPE SR controller and SSA guides) | Shared between arrays? |
|---|---|---|
| Dedicated (default) | Temporarily takes over for a failed drive. When you replace the failed data drive, it must be rebuilt from the data on the spare. | Yes, across multiple arrays on the same controller |
| Auto-replace | The activated spare becomes a permanent member of the array. The original drive bay becomes the spare location, so no second rebuild (copy-back) is needed after you replace the failed drive. | No |
Auto-replace keeps the array running on the spare and saves one copy operation. The trade-off is that data drives move between bays over time, so keep your bay map current.
Activation modes: failure or predictive
The spare activation mode is a controller-wide setting.
- Failure spare activation (default): the spare activates when a member drive fails, and the controller uses the fault tolerance data to regenerate its contents on the spare.
- Predictive spare activation: the spare activates when a member drive reports a predictive failure (SMART). Data is copied to the spare while the volume is still healthy. When the copy completes, the predictive-failure drive is marked failed and ready for replacement. After you install a new drive, the controller restores data from the activated spare to it.
HPE lists these benefits of predictive activation: it is up to four times faster than a typical rebuild, it can recover bad blocks during activation, and it supports all RAID levels, including RAID 0. The SSA GUI guide adds one rule: if you assign auto-replace spares to an array with a RAID 0 logical drive, the spare activation mode must be Predictive Spare Activation.
Step 1 — Set the spare activation mode
In SSA
- Open SSA and the Configure panel, then select the controller in Controller Devices.
- In the Actions panel, click Modify Spare Activation Mode.
- Select Failure Spare Activation or Predictive Spare Activation and click Save.
With ssacli
ssacli ctrl slot=0 modify spareactivationmode=predictive
ssacli ctrl slot=0 modify spareactivationmode=failure
ssacli ctrl slot=0 showThe controller show output includes a Spare Activation Mode line, for example "Activate on drive failure".
Step 2 — Assign spares in SSA
When you create an array, SSA lets you add spares in the same wizard. For an existing array, select the array and use its spare management task.
- In the Configure panel, select the array (or create a new one with Create Array and Create Logical Drive).
- Click Manage Spare Drives (listed as Spare Management in the array task list).
- Select the spare type: Dedicated Spare Drives or Auto-Replace Drives.
- Select the drives that will act as spares and click Save, then Yes to confirm.
- Click Manage Spare Drives again to make more changes, or Finish.
Step 3 — Assign or delete spares in UEFI System Utilities
- Press F9 during POST and open System Configuration > (controller) > Array Configuration > Manage Arrays > (array) > Manage Spare Drives.
- Select Assign Dedicated Spare or Assign Auto Replace Spare.
- Select the drive. Only drives that meet the spare requirements are listed.
To remove a spare, go to Manage Spare Drives > Delete Spare Drives, select the spare and click Delete Spare Drives.
Step 4 — Manage spares with ssacli
The target is an array (or a logical drive if the array has only one). Drive IDs are port:box:bay as shown by ssacli ctrl slot=0 pd all show.
ssacli ctrl slot=0 array A add spares=1I:1:5
ssacli ctrl slot=0 array A add spares=1I:1:5 sparetype=autoreplace
ssacli ctrl slot=0 array all add spares=1I:1:5,1I:1:6
ssacli ctrl slot=0 array A remove spares=1I:1:5
ssacli ctrl slot=0 show configIf you give a drive range, drives that do not meet the spare criteria are skipped. The forced parameter suppresses warning prompts. In SSA Scripting, the matching options are OnlineSpareMode (Dedicated or AutoReplace) and OnlineSpare.
Note: Moving or replacing an array (assigning it to different physical drives) automatically removes its spare drives. Assign the spares again after the move.
Step 5 — Verify and test your process
- Run
ssacli ctrl slot=0 show configand confirm each spare is listed under the right array. - Check the controller's Spare Activation Mode matches your decision.
- Keep a bay map of data drives and spares, especially with auto-replace spares.
- Make sure iLO alerts are configured so a spare activation is noticed; see iLO alerts.
When a spare activates, replace the failed or predictive-failure drive as described in Replace a failed drive on HPE Smart Array.
Frequently asked questions
Can one spare protect several arrays?
Yes, if it is a dedicated spare: HPE describes it as shareable across multiple arrays on one controller. An auto-replace spare cannot be shared.
Does predictive activation work without a spare?
No. Predictive spare activation copies data to an assigned spare, so the array needs at least one spare for the mode to have an effect.
Checklist
- Spare drive is unassigned, same type and at least as large as the smallest array drive.
- Spare type chosen: dedicated (shareable) or auto-replace (no copy-back).
- Spare activation mode set: failure or predictive.
- Auto-replace on an array with RAID 0 uses predictive activation.
- Spares confirmed in
show configand bay map updated. - Alerts configured so spare activation is noticed.
Comments
0 total · 0 threads