none
Inadequate Health Warning Report

    Question

  • I've been running 2012 R2 Essentials (evaluation) for over 3 months now and all has been fine. However I had my first 'fault' issue to deal with yesterday. I first became aware of the issue when I was unable to access files in one of the storage spaces. When I looked at the dashboard I found that the storage space in question had disappeared from view. I looked at the event viewer and found a number of warnings "Disk x has been surprise removed." In fact I found a number of these warnings going back over time but with no real clues from other events around those times. I have since swapped out a PCIEx2 4 Port SATA card which I had the 4 HDD storage space attached to so I can try and locate the issue. But my main point here is there was no indication in the Health report that a storage space was missing. It wasn't listed at all with any indication of trouble. I would have expected it to be listed with a red X next to it to alert me. I am interested in any experiences with this type of scenario and whether I should monitor the system differently. Cheers.

    Update 1

    So, my 4 x HDD (Mirror) storage Space is stable again after swapping out faulty SATA controller – no more event id 157’s so far. This has led me to look further into the question of possible single points of failure for the storage system and the guidelines for connection of physical disks to controllers and the subsequent selection of storage pool disks to create the storage spaces.

    I don’t believe my setup is anything out of the ordinary – motherboard with 6 x onboard SATA ports to which I connected my SSD (OS) and 4 x 4TB HDD. From this storage pool I created a Parity Storage Space (Fixed provisioning, full capacity) with 4 columns to give me 10.9 TB of usable capacity. I needed another storage space so I added the above mentioned PCIEx2 4 port SATA controller and connected 4 x 4TB HDD to increase the storage pool. I then created a new Mirror Storage space (fixed provisioning, full capacity) with 1 column to give me 7.27 TB of usable capacity.

    All well and good but nowhere in the creation wizards is there any notion of which storage pool disks you should pick that considers the controllers they are attached to.

    Is it irrelevant, or should HDD be connected to as many separate controllers that aligns with the disk resiliency layout (mirror, parity) required?

    For instance, should I reconnect my 4 hdd parity and my 4 hdd mirror storage spaces across 4 separate controller cards and would any future single controller failure not cause a storage space to disappear?

    Would really appreciate some experienced help on this, thanks.



    • Edited by naellis Wednesday, October 11, 2017 11:10 AM
    Tuesday, October 10, 2017 2:45 PM

All replies

  • Hi,

    According to your description, my understanding is that health report does not show any message while event id 157 is logged on event viewer for disk remove.

    There are customer identified option for Health Report, one of them is “Critical errors in the event logs”, it will scan Applications and service logs, errors that are logged in the last 24 hours will be presented in report. Event ID 157 is an warning event and belongs to System log, so, it is not reported.

    Storage section in Health Report lists drives and their capacity, it may be used as a reference for checking storage drive.

    Manage System Health in Windows Server Essentials:
    https://docs.microsoft.com/en-us/windows-server-essentials/manage/manage-system-health-in-windows-server-essentials#BKMK_View

    >Is it irrelevant, or should HDD be connected to as many separate controllers that aligns with the disk resiliency layout (mirror, parity) required?
    As Windows Server system itself, there is no mandatory requirement about controller and hardware storage configuration. Basic requirements are minimum hardware space, compatibility. In order to have an further identification about the hardware reliability, usability, I would recommend you to contact the hardware supporter/official website, they may provide you more detail suggestion.

    Besides, below articles can be considered as a reference.

    Storage Spaces Direct hardware requirements:
    https://docs.microsoft.com/en-us/windows-server/storage/storage-spaces/storage-spaces-direct-hardware-requirements

    Windows Server Catalog:
    https://www.windowsservercatalog.com/

    Best Regards,
    Eve Wang

    Please remember to mark the replies as answers if they help.
    If you have feedback for TechNet Subscriber Support, contact tnmff@microsoft.com.


    Thursday, October 12, 2017 7:42 AM
    Moderator
  • Eve, thanks for the reply.

    I understand that MS is not a hardware manufacturer and why you refer me on to one with this matter. However, I am still not clear why a failed controller card only raised Warning level events and no indication of which storage space was affected in the Health report. Can you please explain?

    Can you also confirm that when using the storage space wizards, the disks listed to select from totally ignores the disk-to-controller relationship (for redundancy) when creating parity, mirror storage spaces? I ask this because surely it is relevant in the disk selections made.

    Does using Powershell commands offer anything better in this regard?

    Many thanks

    Neil

    Thursday, October 12, 2017 9:44 AM