Dell PERC, HPE Smart Array (the SR controllers) and HPE MR controllers all build RAID volumes, but each vendor uses its own words, tools and default behaviour. If your team supports both brands, a mistranslated term can lead to the wrong action, for example clearing drives that only needed importing. This guide maps the terms and tools one to one, using only the vendors' own user guides, and points out the behaviours that differ.
Before you start
- Know which controller you have. For ProLiant Gen11, HPE documents both SR (Smart Array) and MR controllers, and states that MR controllers are managed with HPE MR Storage Administrator, not Smart Storage Administrator.
- Dell content here comes from the PERC 11 user's guide, the PERC 12 unified user's guide (H975, H965, H365, HBA465) and the PERC CLI reference for PERC 9, 10 and 11. HPE content comes from the SR Gen10 and Gen11, MR Gen11, SSA CLI and HPE StorCLI guides.
- For RAID theory, see RAID levels and usable capacity.
Terminology map
HPE's own StorCLI guide includes a "Common Branding and Terminology" table that translates MegaRAID/StorCLI words into HPE words: Drive Group (DG) becomes Array, Virtual Drive (VD) becomes Volume, PD becomes Drive, Hot Spare becomes Spare, and Energy Pack or Supercap becomes HPE Smart Storage Battery. PERC tools use the same style of words as StorCLI. The table below lines the three families up.
| Concept | Dell PERC | HPE SR Smart Array | HPE MR |
|---|---|---|---|
| Group of drives that holds RAID | Disk group (perccli: drive group, DGs=) | Array | Array (StorCLI: Drive Group / DG) |
| Volume the OS sees | Virtual disk (VD) | Logical drive | Logical drive (StorCLI: Virtual Drive / VD) |
| Single drive | Physical disk | Physical drive | Drive (StorCLI: PD) |
| Drives passed straight to the OS | Non-RAID disks | Mixed mode: any drive that is not in a logical drive or a spare is presented to the OS by default | JBOD; the Auto-Config option decides the state of a newly added drive |
| Background media check | Patrol read, plus consistency check | Controller surface scan analysis | Patrol read, plus consistency check |
| Cache backup power | Battery | Energy pack (flash-backed write cache) | Energy pack: HPE Smart Storage Battery or Hybrid Capacitor |
Spare drives: the biggest trap
The word "dedicated" does not mean the same thing on every controller:
- Dell PERC: you choose Assign Global Hot Spare or Assign Dedicated Hot Spare in the HII. In perccli, the
dgsoption "specifies the drive group to which the hot spare drive is dedicated". The PERC 12 guide notes that if all the VDs associated with a dedicated hot spare are removed, it becomes a global hot spare. - HPE SR Smart Array: the SR Gen10 guide defines a dedicated spare as "a spare drive that is shared across multiple arrays within a single RAID controller". The other spare type is auto-replace, where the activated spare becomes a permanent array member and no copy-back is needed. Separately, the spare activation mode is either failure or predictive.
- HPE MR: the MR Gen11 guide defines a dedicated spare as dedicated to one array, and a global spare as one that replaces a failed drive in any array if the drive type is the same and its capacity is equal or larger.
So an HPE SR "dedicated" spare behaves more like what PERC and MR users call global. Check the definition for your controller before you plan spare coverage. More detail: global vs dedicated hot spares.
Drives that carry RAID data from elsewhere
PERC and MR use the term foreign configuration: drives that hold RAID metadata the controller does not own. PERC firmware flags such disks as foreign; you can import or clear them in the HII under Main Menu > Configuration Management > Manage Foreign Configurations, or with perccli. The MR guide describes importing or clearing a foreign configuration with HPE MR Storage Administrator. On HPE SR Smart Array, the drive procedures guide describes moving drives and arrays instead: a complete array can move to another controller, even in another server, if all drives move together, the array has no failed or missing drives, and the controller is not in the middle of an expansion or migration.
Warning: clearing a foreign configuration deletes the RAID configuration held on those drives. On PERC 12, the guide also says that with configured slot behaviour, a drive inserted into a configured slot is rebuilt or copied back automatically "regardless of the data on the drive", which overwrites it. Back up first, and never insert a drive that holds data you need into a slot that belongs to a fault-tolerant virtual disk.
Cache settings
- PERC 11: write policies are Write-back (the default) and Write-through. Forced write-back with no battery keeps write-back even without a battery; Dell recommends a power backup system if you use it. Read policies are Read ahead and No read ahead.
- HPE SR: flash-backed write cache (FBWC) uses flash plus the energy pack. In SSA you set the cache ratio; the SR Gen10 guide says the default gives 90% of the cache to writes. Controllers that do not need an energy pack support a no-battery write cache (NBWC) option.
- HPE MR: write policies are Write Through, Write Back and Always Write Back (if the energy pack is absent, the firmware uses Write Back). Read policies are No Read Ahead (the default) and Read ahead.
For what these policies mean in practice, see write-back vs write-through cache.
Tools for each controller
| Where | Dell PERC | HPE SR Smart Array | HPE MR |
|---|---|---|---|
| Before the OS | HII: press F2, Device Settings (Ctrl+R is not supported on PERC 11) | UEFI System Utilities; Intelligent Provisioning | UEFI System Utilities; Intelligent Provisioning 4.31 or later |
| GUI in the OS | Dell OpenManage Storage Management | Smart Storage Administrator (SSA) | HPE MR Storage Administrator |
| Command line | perccli | ssacli | StorCLI (storcli64) |
| Out of band | iDRAC (Comprehensive Embedded Management) | iLO GUI, Redfish, HPE OneView | iLO GUI, Redfish, HPE OneView 8.4 or later |
Read-only status commands, in the syntax each guide documents:
# Dell PERC (perccli /cx show)
perccli /c0 show
perccli /c0/fall show
# HPE SR Smart Array (<target controller> show config)
ssacli ctrl all show config
# HPE MR (storcli64 /cx show all)
storcli64 /c0 show all
storcli64 /c0/fall showFull command guides: perccli status commands, ssacli commands and HPE MR controllers on Gen11.
Frequently asked questions
Can I manage an HPE MR controller with ssacli?
No. HPE states that MR controllers are managed with HPE MR Storage Administrator and cannot be managed with Smart Storage Administrator. Use StorCLI for the command line.
Does Smart Array have patrol read?
The SR guide calls the equivalent controller surface scan analysis. It checks drives in fault-tolerant logical drives for bad sectors and checks parity consistency in RAID 5 or RAID 6 (ADG).
Checklist
- Controller family confirmed: PERC, HPE SR or HPE MR.
- Spare type chosen using that controller's own definition of dedicated and global.
- Foreign or moved drives handled by import, never by clear, unless the data is backed up and not needed.
- Cache policy and backup power (battery or energy pack) checked.
- The right tool installed: perccli, ssacli or StorCLI.
Skyline can supply, configure and maintain Dell, HPE and Huawei servers under an annual maintenance contract (AMC). Contact us.
Comments
0 total · 0 threads