Drivers & software

Upgrade Requirement:
Critical - HPE requires users update to this version immediately.

  The following issues have been fixed in Smart Array controller firmware
   v3.00:

     * Background surface scan might not repair media errors on 12G SAS
       drives in RAID 1, 5, 10, and 50 logical drives.

     * In some I/O workloads, the file system or application might read
       incorrect data from encrypted RAID 0 (with two or more physical
       drives) or RAID 10/10ADM (with four or more physical drives) logical
       drives.

     * A controller might hang when a Test Unit Ready command fails during
       SSD logical drive Rapid Parity Initialization.

     * During a rebuild and extreme simultaneous host I/O writes, the rebuild
       operation might use old data, resulting in subsequent host reads that
       might return old data.

     * A host might read incorrect data after background parity
       initialization finishes on a logical drive on a single array with
       multiple logical drives.

     * A controller lockup issue might occur during the repair of an
       unrecoverable read error (URE) on a RAID 5 volume.

     * A controller lockup issue might occur during SmartCache error
       handling.

     * A controller lockup issue might occur during controller boot when
       SmartCache is enabled.

     * A file system or application might read incorrect data when a
       coalesced host request encounters a fatal error.

     * A failed drive installed during controller boot is not identified as a
       failed drive.

     * The failed drive status LED is not illuminated when a good drive is
       replaced with a failed drive in a backplane.

     * In configurations with backplanes, a failed drive installed as a
       replacement for a previously configured RAID volume is not identified
       as a failed drive.

     * A rebuild of a logical drive created from SSD drives fails to start
       within 1,200 seconds.

     * A performance drop during 16k to 256k sequential reads with low queue
       depth occurs on a logical drive with cache enabled.

     * A controller lockup issue might occur during a simultaneous host write
       and background cache flush operation.

     * Incorrect connector information is displayed for SATA drives after
       certain physical events in a dual I/O module enclosure configuration.

     * A failed drive that is physically present in a drive bay is not
       reported correctly.

     * Devices might drop out of a configuration after multiple create or
       delete drive zone group commands are issued.

     * The drive status LED does not flash during sanitize operations if
       drives are attached to an expander.

     * Cache disable status is not updated correctly when backup power source
       charging times out.

     * The clear configuration command might not clear the configuration if
       SmartCache is enabled and the system was previously shut down
       ungracefully.

     * The clear configuration command fails if a SmartCache volume has
       multiple UREs within a SmartCache page block size.

     * A controller lockup issue might occur when toggling a path to an
       enclosure in a dual domain configuration.

     * The controller runs in survival mode if a temperature sensor is not
       detected.

     * The clear configuration command might report failures when it
       encounters an installed drive with failed status.

     * A performance drop occurs during sequential workloads.

     * A controller failure might occur when running concurrent I/O to
       unassigned drives and drives in a RAID logical drive.

     * After a reboot, a bad drive might prevent controller discovery or
       operating system startup.

     * A host request is incorrectly returned as Failed when a failed drive
       in a degraded volume is replaced.

     * Modifying the spare configuration under an active spare rebuild causes
       the rebuild process to stop.

     * Incorrect device location reporting occurs when direct-attach drives
       are not attached to a backplane and an enclosure management schema is
       not defined.

     * RAID volume configuration settings are not applied when a disk drive
       transitions from unassigned to RAID usage during runtime.

     * A RAID volume with a failed drive reappears after a clear
       configuration operation completes and the system is rebooted.

     * A controller might hang during a SmartCache flush task that encounters
       multiple unrecoverable media errors (UREs).

     * A response of all zeros is sent for ATA PASSTHROUGH commands, such as
       SMART READ DATA, through Out of Band (OOB) host transport.

 The following enhancements are part of the Smart Array controller firmware
   v3.00:

     * After a firmware upgrade and reboot, the background parity
       initialization process starts for logical drives with incomplete or
       failed background parity initialization status.

     * Previously, background surface scans for data consistency could wait
       up to 14 days between runs. Surface scan now restarts immediately when
       a consistency check fails. When no surface scan issues are found,
       surface scan maintains the 14 day waiting period.

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

REMOVED -  Firmware version 2.65 is NO LONGER AVAILABLE for download. Review HPE Customer Bulletin a00097210en_us for additional information that includes replacement firmware version.

• Prevents a potential data inconsistency from occurring after experiencing an Unrecoverable Read Error (URE) on a drive configured in RAID 1/10/ADM Fault Tolerant Mode.
• Prevents a potential data inconsistency from occurring during initial configuration and operation of a volume configured in RAID 5/6/50/60 Fault Tolerant Mode.

o   For additional information, reference Customer Bulletin a00097210en_us.

IMPORTANT: RAID 0 and Unassigned drives are not affected.

Added support for ESXi 7.0

Upgrade Requirement:
Critical - HPE requires users update to this version immediately.

REMOVED -  Firmware version 2.65 is NO LONGER AVAILABLE for download. Review HPE Customer Bulletin a00097210en_us for additional information that includes replacement firmware version.

• Prevents a potential data inconsistency from occurring after experiencing an Unrecoverable Read Error (URE) on a drive configured in RAID 1/10/ADM Fault Tolerant Mode.
• Prevents a potential data inconsistency from occurring during initial configuration and operation of a volume configured in RAID 5/6/50/60 Fault Tolerant Mode.

o   For additional information regarding Compute and Storage platforms, reference Customer Bulletin a00097210en_us.

• HPE does NOT recommend independently updating to Firmware Versions Newer than 1.99 for HPE Smart Array SR Gen10 Controllers on HPE SimpliVity Gen10 systems.
  For additional information regarding SimpliVity platforms, reference Customer Bulletin a00097566en_us

IMPORTANT: RAID 0 and drives in HBA Mode are not affected.

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

REMOVED -  Firmware version 2.62 is NO LONGER AVAILABLE for download. Review HPE Customer Bulletin a00097210en_us for additional information that includes replacement firmware version.

Fixes the following unexpected behaviors:

• Controller could stop responding with a 0x1A91 message during active path failure when scanning expander indexes.
• Controller could stop responding during I/O processing and Smart Cache flush process if there is an invalid transient message return by the physical drives.
• When predictive spare rebuild is triggered on Smart Cache, the rebuilding process could fail to start.
• Data could become inaccessible when a SATA drive overruns a data transfer
• Controller memory could become unstable when the response to an Identify Device command (IDD) is received from a drive
• Communication between the host system and a RAID5/6 volume could potentially fail after an I/O timeout on one of the drives in the array.
• System could stop responding with a 0x1E30 message when an ATA locked drive is discovered
• If a command is execute during a drive sanitize operation the controller could stop responding.

• Add UBM1 Support
• Support for backup power source charge timeout values from UEFI
• UBM2 backplane support
• Drive hotplug support for UBM backplanes

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

REMOVED -  Firmware version 2.02 is NO LONGER AVAILABLE for download. Review HPE Customer Bulletin a00097210en_us for additional information that includes replacement firmware version.

Fixes the following unexpected behaviors:

• Controller could stop responding with a 0x1A91 message during active path failure when scanning expander indexes.
• Controller could stop responding during I/O processing and Smart Cache flush process if there is an invalid transient message return by the physical drives.
• When predictive spare rebuild is triggered on Smart Cache, the rebuilding process could fail to start.
• Data could become inaccessible when a SATA drive overruns a data transfer
• Controller memory could become unstable when the response to an Identify Device command (IDD) is received from a drive
• Communication between the host system and a RAID5/6 volume could potentially fail after an I/O timeout on one of the drives in the array.
• System could stop responding with a 0x1E30 message when an ATA locked drive is discovered
• If a command is execute during a drive sanitize operation the controller could stop responding.

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

REMOVED -  Firmware version 1.99 is NO LONGER AVAILABLE for download. Review HPE Customer Bulletin a00097210en_us for additional information that includes replacement firmware version..

Fixes the following unexpected behaviors:

• Controller could stop responding with a 0x1A91 message during active path failure when scanning expander indexes.
• Controller could stop responding during I/O processing and Smart Cache flush process if there is an invalid transient message return by the physical drives.
• When predictive spare rebuild is triggered on Smart Cache, the rebuilding process could fail to start.
• Data could become inaccessible when a SATA drive overruns a data transfer
• Controller memory could become unstable when the response to an Identify Device command (IDD) is received from a drive
• Communication between the host system and a RAID5/6 volume could potentially fail after an I/O timeout on one of the drives in the array.
• System could stop responding with a 0x1E30 message when an ATA locked drive is discovered
• If a command is execute during a drive sanitize operation the controller could stop responding.

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

REMOVED -  Firmware version 1.98 is NO LONGER AVAILABLE for download. Review HPE Customer Bulletin a00097210en_us for additional information that includes replacement firmware version..

Fixes the following unexpected behaviors:

• Controller could stop responding with a 0x1A91 message during active path failure when scanning expander indexes.
• Controller could stop responding during I/O processing and Smart Cache flush process if there is an invalid transient message return by the physical drives.
• When predictive spare rebuild is triggered on Smart Cache, the rebuilding process could fail to start.
• Data could become inaccessible when a SATA drive overruns a data transfer
• Controller memory could become unstable when the response to an Identify Device command (IDD) is received from a drive
• Communication between the host system and a RAID5/6 volume could potentially fail after an I/O timeout on one of the drives in the array.
• System could stop responding with a 0x1E30 message when an ATA locked drive is discovered
• If a command is execute during a drive sanitize operation the controller could stop responding.

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

Note: If version 1.65 was previously installed, then it is not necessary to upgrade to version 1.65 (B).

• Added support for VMware vSphere 6.7 OS

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

• System might stop responding at POST and fail to load an OS. This issue is more likely to occur if System BIOS version is 1.40 or later. However, systems running older BIOS versions could experience this issue also.

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

• Connectivity to HPE D6020 disk enclosures might be lost if connected to multiple servers.
• In rare cases, data could become unavailable on a RAID 6 (ADG) volume during a rebuild or parity consistency check.
• Inability to configure unassigned drives if a failed RAID volume was present.
• After multiple reboots, unassigned drives might be erroneously assigned as spares for configured RAID volumes.
• System might continuously reboot after splitting a mirrored volume.
• Volume transformation might fail or not resume after a system reboot.
• Maximum command queue depth for SATA drives was limited.
• Controller might stop responding:
  • During power up device discovery. (POST Lockup 0x1ABD or 0x20227)
  • If multiple completions are received for an IO command. (POST Lockup 0x1BE0)
  • If a drive enclosure is power cycled while IO is running. (POST Lockup 0x1C50 or 0x27006)
  • If a drive enclosure is power cycled while a drive is sanitizing. (POST Lockup 0x1E30)
  • When attempting to modify the controller cache settings while IO is running. (POST Lockup 0x1E30)
  • If a drive failed. (POST Lockup 0x27006)
  • When adding or removing drives from a RAID 5 or RAID 50 volume during a rebuild. (POST Lockup 0xFFFFF001)
  • During POST prior to loading an operating system.
  • When Unrecoverable Read Errors (URE’s) are found on drives.
  • When deleting volumes while a rebuild is occurring and URE’s and found.
  • Following completion of Rapid Parity Initialization (RPI) of a new RAID volume.
  • During a volume transformation.
• A drive undergoing RPI might be presented to the operating system.
• Controller slot ID might be shown incorrectly in management tools.
• Incorrect drive bay numbers might be shown in a dual domain configuration.
• Drive sanitize might report incorrect status information.
• Status of the Smart Storage Battery might be reported incorrectly.
• Incorrect POST message 1831 indicating loss of SmartCache data.
• Incorrect text in POST message 1786.
• Various scenarios where drive LEDs might not be displayed correctly.
• The RESTful management interface might fail when attempting to:
  • Change drive write cache configuration.
  • Configure a new RAID volume.

• Increased SmartCache maximum logical volume size.
• Added support for SmartCache in UEFI configuration tools.
• Added the ability to control drive write cache settings.
• Added the ability to select the drive erase pattern.

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

Note: If version 1.34 was previously installed, then it is not necessary to upgrade to version 1.34 (B).

• Added support for the HPE Smart Array P408e-m Controller.

Upgrade Requirement:
Recommended - HPE recommends users update to this version at their earliest convenience.

• Controller might stop responding while handling a drive error.  (POST Lockup 0x27006)
• Controller might stop responding after a drive hot-removal or hot-insertion.  (POST Lockup codes 0x1EA or 0x15010)
• Controller might stop responding during a volume expansion or rekey.  (POST Lockup code 0x1E30)
• Controller might stop responding while performing heavy writes to a RAID 0 or RAID 1 logical drive that has a flash backed write cache enabled.  (POST Lockup 0x1BC0)
• Various scenarios where drive LEDs might not be displayed correctly.
• Drives might not be detected after a cold boot.
• A hot-inserted drive might be immediately reported as failed.
• A failed or charging HPE Smart Storage Battery might be reported as “missing” in SSA and iLO.
• The RESTful management interface might be unable to detect physical drives or make configuration changes to the controller.
• Option to enable “No Battery Write Cache” in SSA is not available if an HPE Smart Storage Battery is not installed.
• Customer specified read/write controller cache ratio is not honored if “No Battery Write Cache” is enabled.
• Controller cache might be displayed as temporarily disabled following an unexpected shutdown and subsequent reboot.
• Unable to modify the controller cache read/write ratio following an unexpected shutdown and subsequent reboot.
• The controller cache might become disabled when running VMware ESXi 6.0 if the controller and guest OS are configured to use PCI pass-through.
• If multiple logical volumes are present on the same array, and one of the volumes is rebuilding, the others might be incorrectly marked as failed.
• A drive at index 0 might incorrectly display as rebuilding if a logical volume on the controller is currently rebuilding.
• System might hang during power-up or appear unresponsive during runtime if a logical volume composed of SSD’s is created or rebuilding.
• Negotiated link rate for drives might not be correct.
• Storage enclosure information for direct attached drives might not be returned correctly in Windows.
• The HPE SmartCache feature was not available by default on the P816i-a SR Gen10 controller.

• Support for ATA Sanitize Freeze/Antifreeze Lock.
• Ability to select boot controller in Legacy BIOS mode with multiple controllers installed.
• Updated Hard Drive Carrier LED behavior for the HPE ProLiant XL450 Gen10 Server.
• Ability to set rebuild priorities via the RESTful management interface.
• Supported settings – High, MediumHigh, Medium, Low