Launching soon: The world's first vendor agnostic bug scrubLearn more & join waitlist
OPERATIONAL DEFECT DATABASE
...
...
A rare issue has been identified in the HPE ProLiant for Microsoft Azure Stack Hub Gen10 solution. In certain scenarios involving multiple concurrent SRB_FUNCTION_IO_CONTROL function calls (I/O control requests), to retrieve drive firmware information (IOCTL_FIRMWARE_FUNCTION_GET_INFO ) within Azure Stack Hub storage, excessive I/O request response times may occur, particularly in environments with 12 or more cluster nodes and a significant volume of I/O control requests. This occurs when using a Smart Array SR Gen10 controller running SmartPQI Driver for Windows v1010.42.0.1020, which is part of HPE ProLiant for Microsoft Azure Stack Hub Solution Update Bundle 4.2.0 and 4.3.0. These prolonged response times have the potential to generate a cascading effect manifested as Storage Spaces Direct (S2D) I/O timeouts, potentially resulting in cluster node outages. When this occurs, a Storport Event 500 will log a timeout on a SrbFunction 2 command, followed by a reset and then all the drives being marked unresponsive. Microsoft-Windows-Storage-Storport-Operational.EVTX Provider = Microsoft-Windows-StorPort Event ID = 500 Level = 2 (Error) Example description: "A request timed out for Storport Device (Port = 2, Path = 2, Target = 2, Lun = 14). Corresponding Class Disk Device Guid is { 623204f3-7a31-5523-a7ac-36f3fe830302}."
Any of the following HPE Smart Array SR Gen10 controllers running SmartPQI Driver for Windows v1010.42.0.1020 or later in HPE ProLiant for Microsoft Azure Stack Hub Gen10 solution are affected by this issue: HPE Smart Array P408i-p SR Gen10 Controller HPE Smart Array P408e-p SR Gen10 Controller HPE Smart Array P408i-a SR Gen10 Controller HPE Smart Array E208i-p SR Gen10 Controller HPE Smart Array E208e-p SR Gen10 Controller HPE Smart Array E208i-a SR Gen10 Controller HPE Smart Array P816i-a SR Gen10 Controller
This has been resolved with HPE ProLiant for Microsoft Azure Stack Hub Solution Update Bundle version 4.4.0.6, available here . As a workaround for earlier versions, perform the following operation on each cluster node: Create a new REG_BINARY type Microsoft Windows registry setting named "DisableSendingBmicSDUS" under HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\SmartPqi\Parameters\Device and set the value to "hex:01." This registry setting will allow the Smart Array Gen10 controller to use its cached drive firmware version and a SCSI INQUIRY will not be sent to the drives for every IOCTL_FIRMWARE_FUNCTION_GET_INFO request. CAUTION: This document contains information about how to modify the registry. Use the Registry Editor at your own risk! The registry should be backed up before modifying it. If the Registry Editor is used incorrectly, serious issues could occur that may require a reinstallation of the operating system. For more information about how to back up, restore, and modify the registry, see Microsoft Knowledge Base Article 256986, titled Windows registry information for advanced users. Click here to view Microsoft Knowledge Base Article 256986 . Disclaimer: the link above will take you outside the HPE website. HPE is not responsible for content outside of its domain. Revision History Document Version Release Date Details 2 June 18, 2024 Updated Resolution with permanent fix. 1 February 7, 2024 Original Document Release.