Troubleshooting EMC systems and hypervisors

When a valid description is typed in the Storage Pool Description field, sometimes the description is not displayed in the column on the Storage Pool tab.

This is a known problem. The description display is for more information and is not critical for system operations.

• ESI fails to connect to Windows clusters or failover clusters.

• ESI cannot create, delete, or disconnect a cluster disk.

• Healthy clusters are an ESI requirement. Confirm that the clusters are healthy.

• Cluster resource dependencies are not removed before you delete or disconnect a cluster disk. Before deleting or disconnecting a cluster disk, remove disk resource dependencies.

Ping system timeout fails.

Set the timeout value according to your network status in ESI.

You are unable to connect to a Windows server.

• Check that firewall rules are enabled on both controller and controllee hosts.

• Check that remote PowerShell is enabled on both controller and controllee hosts.

• Check that DNS is configured correctly.

Rescan fails to find a LUN.

• Check that the FC Zoning is configured correctly.

• Check that the iSCSI initiator is logged in to the target port.

• Check that the multipath software is configured correctly.

• Storage-related data does not display in the host view, cluster view, or SharePoint view.

• The Storage pools in the create disk wizard do not load and the wizard is blank.

• ESI shows storage related information for a disk or shared folder only when the corresponding storage systems are registered. Verify that the corresponding storage systems are registered with ESI (under Storage Systems). If not, register these storage systems.

• The storage system may be taking longer than the default time-out value of 60 seconds. Set a higher time-out value (in seconds) in the following Registry key:

Key: HKLM\SOFTWARE\EMC\WSI\Config

DWORD Value Name: DefaultOperationTimeout

• When a storage system is added to ESI, it is not listed in PowerShell.

• When a storage system is added in PowerShell, it is not listed in ESI.

Log out of the other application and then reopen it to refresh the list.

"Retrieval of host disks failed: Operation RefreshDisks failed on Vds proxy..." error appears.

When the VDS load operation uses the target host FQDN, this name resolution error occurs. To resolve this error, you must configure the Windows hosts file on the ESI controller host system with both the short (NETBIOS) and long (FQDN) host names.

When importing storage with the access control file, some storage does not display.

Before importing the storage access control file, confirm that the ESI host has the respective storage adapters installed on it.

In ESI, right-clicking a tree node does not display all the menu options.

This is the default behavior of the Microsoft Management Console Framework. Select an item, and then right-click to display the menu options.

• When creating a disk, ESI fails or creates the incorrect volume size.

• If the file system type is FAT32, provisioning a storage volume fails.

If you create a host disk and ESI fails or creates the incorrect volume size, you may need to increase the disk size. For example, when creating a 16 KB cluster unit with a FAT32 file system type, the minimum disk size requirement is 1,029 MB.

A FAT32 volume must contain a minimum of 65,527 clusters. You cannot increase the cluster size on a volume that uses the FAT32 file system so that it contains fewer than 65,527 clusters. Limitations of the FAT32 File System in Windows XP on the Microsoft Support website has more details at:

http://support.microsoft.com/kb/314463

Sometimes the smaller clusters can fail because the required 65,527 cluster minimum is not met.

You are unable to view the mounted shared folders on the target host for file storage systems.

Log out and then log back in to the server. This reinitializes the shared folder view. After logging back on, you can view the shared folder changes on the target host. The mapping between drives and shared folders is specific to each user account, unlike the LUN or disk mounting that does not vary between accounts. Every user account can have different drive mappings for the same drive letter.

When you specify both the -ID and -ConcreteLun parameters together with the Get-EmcLun ESI PowerShell cmdlet, an AmbiguousParameterSet exception error occurs.

The -ConcreteLun switch parameter specifies the Concrete LUN type as an optional parameter, through which only Concrete LUNs can be filtered out. Do not use this parameter with the -ID parameter. This parameter supports only block storage systems.

When you create a virtual hard disk or passthrough SCSI disk for a virtual machine, no IDE controllers are listed.

ESI does not support IDE-based disks. Use hypervisors to create IDE-based disks.

For hypervisors, cannot create new SCSI controllers for virtual machines.

ESI does not support creating new SCSI controllers for creating disks for virtual machines. Use the hypervisor to create new SCSI controllers. Then use ESI to create and attach the disks for virtual machines with these existing SCSI controllers.

For hypervisors, the Connect to Host action leads to the following error message: "Can't retrieve IP from MAC address:…" or the "host name is empty."

This error can occur for a number of reasons. Confirm the following to avoid this error:

• The virtual machine is a part of a reachable domain.

• The supported Windows operating system is installed on that virtual machine.

• The IP of that virtual machine is configured correctly.

• The ESI-mandatory firewall settings are configured correctly if you want to manage the virtual machine in ESI.

For VMware, you cannot expand RDM disks that are attached with virtual compatibility mode.

For VMware systems, ESI does not support RDM disks that are created with virtual compatibility mode. Use ESI to provision disks without the virtual compatibility mode setting.

For a VMware ESX host that is connected with ESI to a virtual machine, when expanding a file-based disk on the virtual machine, the operation fails with an "Access to resource settings on the host is restricted to the server that is managing it" error message.

This is a VMware restriction for expanding file-based disks, which requires the operation to only succeed when only the vCenter host is connected and the ESX host is disconnected from ESI. Try removing the ESX host and adding the vCenter hypervisor again in ESI. Then try the expand operation again.

For a XenServer, cannot add or remove a virtual disk on a running virtual machine.

Occurs when xs tools is not installed on XenServer.

XenServer requires that xs tools be installed on the virtual machine before you can add or remove virtual disks while the virtual machine is running. You can either install xs tools or shut down the virtual machine before adding or removing disks.

For XenServers, the operation fails when attempting to remove a XenServer storage repository.

Failure occurs when you create a XenServer storage repository on a LUN and then mask the LUN without deleting the storage repository.

When you click Remove Storage Repository, XenServer cannot discover the LUN mask immediately, even if the page is refreshed, and the operation fails.

Try completely removing a storage repository as described in Forgetting Citrix XenServer storage repositories.

And then when you create a LUN for a storage repository, follow the steps in Creating Citrix XenServer storage repositories.

For XenServers, when you create or delete a virtual disk for a storage repository, the free capacity of the storage repository is not updated immediately.

In the ESI Window, go to the Storage Repositories tab. In the Actions pane, click Refresh to update the storage repository disk information.

For CX4 systems and XenServer hypervisors, ESI displays the incorrect LUN size for thin provisioning.

This is a known problem with CX4 systems and XenServers. XenServer assigns a thin LUN from ESI as smaller than the assigned size. To avoid this error, create thick LUNs for XenServer storage repositories.

For CX4 and VNX systems, the local administrator account status appears as Offline.

This occurs when a CX4 or VNX storage array is added with a global account without administrator rights or a local administrator account, and the status is shown as Offline. VNX and CX4 systems require administrator accounts with a global scope.

For CX4 and VNX systems, unable to connect a host disk.

If the CX4 or VNX storage array does not have Access Logix enabled, the host disk connection fails. CX4 and VNX block storage systems must have Access Logix enabled on a storage array before you connect a host disk on a Windows host.

For VNX systems, advanced snapshot LUNs can be viewed on the LUNs tab in the ESI GUI, but are not listed in ESI PowerShell.

You can view advanced snapshot LUNs in ESI GUI. Even though the snapshots exist, you cannot currently view the list with the Get-EmcSnapshotLuns cmdlet.

For VNX-CIFS systems, a time-out error message occurs.

In ESI version 2.0, the default File Port Number was 443. In ESI version 3.0 and later, the default is 22.

If you upgraded to version 3.0 from a previous release, you may have set up an incorrect File Port Number. To change the port, remove the VNX-CIFS storage system and then add it again with the correct default port number of 22.

For VNX-CIFS systems, a "Host key was not found in the repository: <IP Address>" or a "Host key has been changed" error message appears.

The Add Host Key If Missing and Replace Host Key If Changed settings are new as of the 3.0 release. You might need to reset these security validation settings for VNX-CIFs systems.

Or if the host key was recently changed on the VNX file storage system, reset the Registry key with ESI.

To reset the Registry key settings:

1. Remove the VNX-CIFS storage system.

2. Add the system again and update the Registry key information:

• • Select the Replace Host Key If Changed checkbox and click Test Connection. ESI will read and update to the new host key information during the connection test.

• Clear that checkbox and then click Add to add the system.

This maintains security, yet allows ESI to connect to the system and update the cached host key in ESI.

For VNX-CIFS systems, the connection fails.

ESI requires the correct control station username and password is set up for the VNX-CIFS connection. ESI does not support root for the VNX-CIFS control station connection. If necessary, you can use nasadmin or equivalent administrator username and password to connect to VNX CIFS systems.

For VNXe and Hyper-V, when you use the New-EmcLun cmdlet to create a LUN for Hyper-V, a "LUN does not exist" error occurs.

This occurs in older versions of VNXe. To resolve this problem, update your VNXe to version 2.3.1.20364.

For VNXe, a snapshot cannot be promoted.

If the resource has no access to the snapshot, a snapshot cannot be promoted for a host. Set up host access to the snapshot, and then the snapshot can be promoted for that host.

When adding a VNXe, the "System cannot find the file specified" error appears when you click Test Connection.

VNXe Unisphere CLI is not available on the controller host where ESI is running. Download and install Unisphere CLI from the EMC Online Support website.

After installing VNXe Unisphere CLI on the controller host, retry adding the system.

PowerShell scripts from earlier versions of ESI that use the Connect-EmcSystem PowerShell cmdlet fail with the "Parameter ConnectionName is needed to create the Host System Object" error.

For ConnectEmcSystem, use ConnectionName instead of HostSystemName for the Creation parameter.

• The Set-EmcLunAccess PowerShell cmdlet fails.

• An unmasking operation fails.

Before unmasking a LUN or using the cmdlet, ESI requires that you register all iSCSI and FC HBAs or host initiators used to unmask LUNs on the storage system.