로그인
 
홈라이브러리학습다운로드지원커뮤니티포럼
Resources for IT Professionals >
> 
> 
Cluster Shared Volume does not mount properly after a restart of both nodes?

Question Cluster Shared Volume does not mount properly after a restart of both nodes?

  • 2012년 8월 5일 일요일 오후 10:27
     
     
    로그인하여 투표
    0

    Hardware:

    2x Dell PowerEdge R510

    1x Dell PowerVault MD3220i

    Background:

    Servers connected to SAN via ISCSI through 2 host ports each via a Dell PowerConnect Layer 2 switch.

    Running two Nodes with Hyper-V Core installs and Failover Feature Enabled.

    Validated and setup Failover Cluster for server.

    Failover works (IE. VMs will be moved to Node 2 when Node 1 goes offline)

    Migrations will work (live, move and quick).

    Problem:

    When both servers where recently shutdown for power maintenance in the building, after being turned back on both servers failed to mount the Cluster Shared Volume properly.

    The Shared Volume is showing Online in the FailOver manager, but in iscsicpl.exe the mount point for the volume is not set to c:\clusterstorage\volume1 but instead it /?/volume/{GUID}/

    When trying to direct to c:\clusterstorage\volume1 it is listed but freezes when trying to access it.

    If you remove the volume from shared it mounts properly.

    TL:DR

    Cluster shared volume doesn't mount after simultaneous restart of servers, only show the mount point as /?/volume/{GUID}/ cannot remount as shared.

     

모든 응답

  • 2012년 8월 6일 월요일 오전 1:45
    중재자
     
     
    로그인하여 투표
    0

    Hi,

    but in iscsicpl.exe the mount point for the volume is not set to c:\clusterstorage\volume1 but instead it /?/volume/{GUID}/
    >> Actually, Cluster Shared Volumes(CSV) is not configured in iscsicpl.exe(iSCSI Initiator console), you need to do it in Failover Cluster console.

    By the way, what did you mean by “If you remove the volume from shared it mounts properly.”?

    In additoin, did you get any errors in Event Viewer about Hyper-V and Failover Cluster?

     
  • 2012년 8월 7일 화요일 오후 7:46
     
     
    로그인하여 투표
    0

    Hi,

    but in iscsicpl.exe the mount point for the volume is not set to c:\clusterstorage\volume1 but instead it /?/volume/{GUID}/
    >> Actually, Cluster Shared Volumes(CSV) is not configured in iscsicpl.exe(iSCSI Initiator console), you need to do it in Failover Cluster console.

    By the way, what did you mean by “If you remove the volume from shared it mounts properly.”?

    In additoin, did you get any errors in Event Viewer about Hyper-V and Failover Cluster?

    I understand that the CSVs are set up via the Failover Management, yes, but should the mountpoint in iscsicpl not reference the actual mountpoint of the volume (ie. if it's mounted to c:\clusterstorage\volume1).

    Situation 1:

    CSV configured > Mountpoint (according to Failover Console) c:\ClusterStorage\Volume1 > unabled to access volume (Mountpoint according to iscsicpl is /?/volume/{GUID}/)

    Situation 2:

    CSV configuration is removed > Mountpoint (according to Failover Console) E: drive > able to access volume (Mountpoint according to iscsicpl is E:)

    There would be no errors (besides the power/quorum warnings from after the restart).

     
  • 2012년 8월 7일 화요일 오후 9:13
     
     
    로그인하여 투표
    0
    Hi, the mount points will always be the random guid. thats perfectly fine. Anything in the cluster eventlog or eventlog on the nodes?

    Jaysam Thanki

     
  • 2012년 8월 7일 화요일 오후 9:50
     
     
    로그인하여 투표
    0
    Hi, the mount points will always be the random guid. thats perfectly fine. Anything in the cluster eventlog or eventlog on the nodes?

    Jaysam Thanki

    Ahh, right, okay, hmm.

    First few events from after the reboot (from node 1) that are notable. The last one looks especially suspicious.

    Log Name:      System
    Source:        iScsiPrt
    Date:          8/5/2012 9:33:15 AM
    Event ID:      5
    Task Category: None
    Level:         Error
    Keywords:      Classic
    User:          N/A
    Computer:      HVS1.domain.local
    Description:
    Failed to setup initiator portal. Error status is given in the dump data.

    Log Name:      System
    Source:        MSiSCSI
    Date:          8/5/2012 9:33:43 AM
    Event ID:      113
    Task Category: None
    Level:         Warning
    Keywords:      Classic
    User:          N/A
    Computer:      HVS1.domain.local
    Description:
    iSCSI discovery via SendTargets failed with error code 0xefff0003 to target portal *192.168.31.3 0003260 Root\ISCSIPRT\0000_0 192.168.31.1.

    Log Name:      System
    Source:        iScsiPrt
    Date:          8/5/2012 9:33:43 AM
    Event ID:      70
    Task Category: None
    Level:         Error
    Keywords:      Classic
    User:          N/A
    Computer:      HVS1.domain.local
    Description:
    Error occurred when processing iSCSI logon request. The request was not retried. Error status is given in the dump data.

    Log Name:      System

    Source:        MSiSCSI
    Date:          8/5/2012 9:33:52 AM
    Event ID:      121
    Task Category: None
    Level:         Warning
    Keywords:      Classic
    User:          N/A
    Computer:      HVS1.domain.local
    Description:
    The firewall exception to allow Internet Storage Name Server (iSNS) client functionality is not enabled. iSNS client functionality is not available.

    Log Name:      System
    Source:        MSiSCSI
    Date:          8/5/2012 9:35:25 AM
    Event ID:      103
    Task Category: None
    Level:         Error
    Keywords:      Classic
    User:          N/A
    Computer:      HVS1.domain.local
    Description:
    Timeout waiting for iSCSI persistently bound volumes. If there are any services or applications that use information stored on these volumes then they may not start or may report errors.