none
Get-StorageEnclosure Command concerns RRS feed

  • Question

  • We recently purchase 4 INTEL® JBOD2224S2DP units and 2 Dell Poweredge 520 units with 2 LSI cards (SAS9201-16e) in each server with the intention of using the JBOD units with Windows Storage Spaces.

    ConnectionsUsing the diagram above is how we connected the systems at first.  We where able to see all the hard drives within storage spaces and we where able to create a cluster and utilize the storage pools.When we tried to do a more advanced configuration using Powershell, we are getting a response from Windows saying that the Enclosure is in an unhealthy state and we can only see one enclosure ,

    • All firmware and BIOS updates have been applied. 
    • We are using Windows 2012 R2 Core, all updates as of Sept 5 2014.  hotfix KB2913766.
    • We installed LSI drivers vs windows built in drivers.
    • We installed the management console for the LSI card which can see the enclosures themselves and also blink the hard drives.
    • We installed hotfix KB2913766.
    • We tried connecting just 2 JBOD units alone into one server
    • We tried removing a Single LSI card
    • We tried using another DELL SAS card with both LSI cards removed
    • We have tried connecting one jbod unit at a time, which we have noticed that the last JBOD unit plugged in is the one that the Windows system pulls information from, One thing of note, the Unique Id is the same regardless of which unit was plugged in.

    We contacted INTEL about this but they never give us a straight answer, and even tell us that the LSI card is not verified by intel and they recommend actual RAID cards which wont work with Storage Spaces, so when I asked if they have someone with experience with storage spaces they avoided answering  the question like a politician.

    Is there a configuration step im missing here?

    Friday, September 5, 2014 4:19 PM

All replies

  • EDIT: My apologies. Got the wiring diagram backwards initially.

    First: HBA 0 on both servers should be connected to Expander A.

    Second: HBA 1 on both servers should be connected to Expander C.

    This will give a redundant path two the drives themselves based on your above drawing.

    Also, make sure all firmware is up to date in the HBAs and the JBODs.

    Intel has a Technical Advisory based on the Same ID Issue here: TA-1042: Storage server system can show incorrect system enclosure IDs.

    Intel also has a firmware update in the works to hopefully correct the problem.


    Philip Elder Microsoft Cluster MVP Blog: http://blog.mpecsinc.ca


    • Edited by Philip ElderMVP Monday, September 8, 2014 5:08 PM Misread the diagram initially.
    Monday, September 8, 2014 3:27 PM
  • Thanks for the reply. I will try the configuration you mentioned, right now however redundancy is not my primary concern.  Even with one unit connected the command is showing that its unhealthy without redundancy in mind that command should show all enclosures.

    All firmwares and hotfixes where applied from Intel, MS and LSI.  I have a support ticket open with MS now, as Intel is saying that its not an issue with the hardware.

    Monday, September 8, 2014 7:02 PM
  • Do you have the following hotfix installed on all SOFS Nodes: KB2913766: Hotfix improves storage enclosure management for Storage Spaces in Windows 8.1 and Windows Server 2012 R2?

    Please read the linked to Intel Technical Advisory. There is a known problem with the Enclosure IDs.

    The following is a copy/paste of our lab setup:

    Get-StorageEnclosure

    ObjectId                           : {1}\\LAB-HVNODE-01\root/Microsoft/Windows/Storage/Providers_v2\SPACES_StorageEnclo
                                         sure.ObjectId="{f95198f4-e2c7-11e3-80b5-806e6f6e6963}:SE:{0b45bd24-a599-a438-b0fa-
                                         bce348727789}"
    PassThroughClass                   :
    PassThroughIds                     :
    PassThroughNamespace               :
    PassThroughServer                  :
    UniqueId                           : 5001E67800FFFFFF
    CurrentSensorOperationalStatus     :
    DeviceId                           : {0b45bd24-a599-a438-b0fa-bce348727789}
    FanOperationalStatus               : {OK, OK, OK}
    FirmwareVersion                    : 0d00
    FriendlyName                       : StorageEnclosure2
    HealthStatus                       : Unhealthy
    IOControllerOperationalStatus      :
    Manufacturer                       : INTEL  
    Model                              : RES2CV360-R    
    NumberOfSlots                      : 24
    PowerSupplyOperationalStatus       : {OK, OK}
    SerialNumber                       :
    TemperatureSensorOperationalStatus : OK
    VoltageSensorOperationalStatus     : {OK, OK, OK}
    PSComputerName                     :

    ObjectId                           : {1}\\LAB-HVNODE-01\root/Microsoft/Windows/Storage/Providers_v2\SPACES_StorageEnclo
                                         sure.ObjectId="{f95198f4-e2c7-11e3-80b5-806e6f6e6963}:SE:{25bb8091-fb81-e5aa-543a-
                                         34046203b0d1}"
    PassThroughClass                   :
    PassThroughIds                     :
    PassThroughNamespace               :
    PassThroughServer                  :
    UniqueId                           : 5001E676CFD13FFD
    CurrentSensorOperationalStatus     :
    DeviceId                           : {25bb8091-fb81-e5aa-543a-34046203b0d1}
    FanOperationalStatus               : {OK, OK, OK}
    FirmwareVersion                    : 0d00
    FriendlyName                       : StorageEnclosure0
    HealthStatus                       : Unhealthy
    IOControllerOperationalStatus      :
    Manufacturer                       : INTEL  
    Model                              : RES2CV360-R    
    NumberOfSlots                      : 24
    PowerSupplyOperationalStatus       : {OK, OK}
    SerialNumber                       :
    TemperatureSensorOperationalStatus : {OK, OK, OK, OK}
    VoltageSensorOperationalStatus     : {OK, OK, OK}
    PSComputerName                     :

    ObjectId                           : {1}\\LAB-HVNODE-01\root/Microsoft/Windows/Storage/Providers_v2\SPACES_StorageEnclo
                                         sure.ObjectId="{f95198f4-e2c7-11e3-80b5-806e6f6e6963}:SE:{73dfb05a-d5a6-7063-d1ee-
                                         838fef639fd2}"
    PassThroughClass                   :
    PassThroughIds                     :
    PassThroughNamespace               :
    PassThroughServer                  :
    UniqueId                           : 5001E676CFD48FFD
    CurrentSensorOperationalStatus     :
    DeviceId                           : {73dfb05a-d5a6-7063-d1ee-838fef639fd2}
    FanOperationalStatus               : {OK, OK, OK}
    FirmwareVersion                    : 0d00
    FriendlyName                       : StorageEnclosure1
    HealthStatus                       : Unhealthy
    IOControllerOperationalStatus      :
    Manufacturer                       : INTEL  
    Model                              : RES2CV360-R    
    NumberOfSlots                      : 24
    PowerSupplyOperationalStatus       : {OK, OK}
    SerialNumber                       :
    TemperatureSensorOperationalStatus : {OK, OK, OK, OK}
    VoltageSensorOperationalStatus     : {OK, OK, OK}
    PSComputerName                     :


    Philip Elder Microsoft Cluster MVP Blog: http://blog.mpecsinc.ca

    Monday, September 8, 2014 7:43 PM
  • Yep we have the patch from Intel and MS applied, in your case your getting 3 different enclosures I'm only getting one, even with the configuration suggested.

    Also your read out also shows as unhealthy, that's the status im worried about.

    • Edited by ITNCC Tuesday, September 9, 2014 1:11 PM
    Tuesday, September 9, 2014 12:42 PM
  • That has been the case for most folks that I have worked with.

    Apparently there is a new firmware in the works that is supposed to finally address this problem. I suggest opening a case with Intel if you have not already done so. If you have ping me off-list with the support case ID and I will forward that on to our Intel contacts.

    EDIT: I just heard back from one of our contacts with the same problem that they received the new firmware and it did indeed fix the Enclosure ID problem. Please reach out to Intel.


    Philip Elder Microsoft Cluster MVP Blog: http://blog.mpecsinc.ca


    • Edited by Philip ElderMVP Tuesday, September 9, 2014 4:12 PM Added the firmware update note.
    Tuesday, September 9, 2014 3:43 PM
  • I will double check on the firmware, but I'm pretty sure we applied the newest one, Ill get a screen shot of the version i have. I even had a Intel Tech on the phone with me while I updated it.  Ive sent you a message through your blog with my case number.

    Thanks again for the assistance.

    Edit:This is a screen shot of what I have the JBOD units upgraded to


    • Edited by ITNCC Tuesday, September 9, 2014 6:38 PM attach pic
    Tuesday, September 9, 2014 4:45 PM
  • Hi, I've received your case ID.

    On the blog is an e-mail link. Please ping me via that as I may have further info on what needs to be done.


    Philip Elder Microsoft Cluster MVP Blog: http://blog.mpecsinc.ca

    Tuesday, September 9, 2014 7:06 PM
  • Thanks again for the assistance.  The firmware that was recently released, 00.13.00.09 did in fact correct the issue with the hardware IDs.

    Should also be noted that the firmware updates on a per expansion card basis.  Should be obvious but when I did the initial flash of the firm ware I had not realized that the ports i had connected where not plugged into both adapters, so one expander on some of the units had the original firmware, 255.255.255.255, while others had the 00.13.00.08 for the second expander.

    The only problem remaining is that the command still shows enclosures as Unhealthy which is inline with what you currently have.

    Monday, September 15, 2014 12:30 PM
  • Thanks again for the assistance.  The firmware that was recently released, 00.13.00.09 did in fact correct the issue with the hardware IDs.

    Should also be noted that the firmware updates on a per expansion card basis.  Should be obvious but when I did the initial flash of the firm ware I had not realized that the ports i had connected where not plugged into both adapters, so one expander on some of the units had the original firmware, 255.255.255.255, while others had the 00.13.00.08 for the second expander.

    The only problem remaining is that the command still shows enclosures as Unhealthy which is inline with what you currently have.

    I suggest having the second set of cables connected to port C on the JBODs. Both expanders would then be flashed.

    No changes need be made at the OS level. MPIO will adapt.

    Now that the Enclosure ID shows correct Enclosure Awareness can be enabled and resilience tested.


    Philip Elder Microsoft Cluster MVP Blog: http://blog.mpecsinc.ca

    Monday, September 15, 2014 1:46 PM
  • Yep that's whats showing now, should we be concerned about the Unhealthy status? I'm still awaiting Intel's reply on that issue.

    Monday, September 15, 2014 8:18 PM
  • I'm not sure. Our goal is to start enclosure resilience testing in the next week or so.

    Hopefully that unhealthy status does not trigger a full-stop for the entire setup.

    Hopefully Intel can give us some direction because there are a number of folks in the same boat.


    Philip Elder Microsoft Cluster MVP Blog: http://blog.mpecsinc.ca

    Monday, September 15, 2014 9:18 PM
  • Sounds good, out of curiosity, what are you doing to perform your resilience testing?  we are using a few scripts to test IOPs using different RAID setups, and the usual pulling of cables, power supplies and hard drives to see how the system reacts.

    Wednesday, September 17, 2014 7:24 PM