Installation
This chapter describes how to identify and resolve installation problems.
Installer Application Fails
The following are symptoms, possible causes, and solutions for when the installer application fails.
|
|
|
The Hyper-V host is not visible in the installer application UI. |
The hosts management adapter is already associated with a Virtual switch. |
1. Launch the Microsoft System Center Virtual Machine Manager (SCVMM) user interface (UI). 2. Choose the host in the left hand pane. 3. Choose Virtual Switches. 4. From the list choose the switch that has the host management adapter and choose Delete from the toolbar. 5. Verify that the job is complete by checking in the Jobs section. 6. Run the installer application again. |
The host might be down. |
Refresh the host, see “Host is in the Not Responding State in the Microsoft SCVMM” section |
The VSM VM is created but it is not responding. |
- The upstream switch is not configured properly (VLAN is not allowed).
- The Proxy setting is enabled on the Windows server where the Microsoft SCVMM is installed.
|
1. Ping the VSM IP – If you can reach the VSM then check that the proxy settings are disabled on the Windows server where Microsoft SCVMM is installed. – If you are not able to reach the VSM then check the upstream switch configuration. 2. Run the installer application again. |
Rollback message is displayed. |
The installer application is unable to continue. The reason for the error is displayed in the rollback message. |
1. Copy the rollback message and choose to rollback automatically or manually: a. Choose Yes to rollback automatically.
Note If an error occurs during the rollback you must manually clean up the installation.
b. Choose No and you must manually clean up the installation. 2. Using the information from the rollback message resolve the error. 3. Run the installer application again. |
Microsoft Hyper-V manager is missing. |
The host was added to Microsoft SCVMM without enabling the HyperV role manually. |
1. Install the Microsoft Hyper-V manager by entering the following command: start /w ocsetup Microsoft-Hyper-V-Managment-Clients |
Host is in the Not Responding State in the Microsoft SCVMM
You can refresh the host that is in the Not Responding state.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Choose the server that is in the Not Responding state.
Step 3 Refresh the host.
Installation Failure When Microsoft SCVMM Fails to Resolve Host Names
The Microsoft SCVMM might fail to resolve the host names of the managed Cisco Nexus 1000V for Microsft Hyper-V servers. This may result in the failure of pushing Cisco Nexus 1000V for Microsoft Hyper-V VEM MSI to the Microsoft SCVMM server hosts from the Microsoft SCVMM server.
Any host side operation may fail when DNS is not resolved resulting in:
- Refresh failure of the host from Microsoft SCVMM
- Failure to create a Cisco Nexus 1000V logical switch on the host
Step 1 Launch the Microsoft SCVMM UI.
Step 2 At the command prompt, enter ping hostname , where hostname is the name of the DNS host in question.
Step 3 Enter winrm id - r < hostname >.
Step 4 Repeat Step 2and Step 3 from the host and replace hostname with the name of the Microsoft SCVMM DNS server.
Step 5 If there is more than one DNS server associated with the host, make sure the management NIC contains only the DNS server pointing to the AD.
Step 6 Using your browser, navigate to Tools > Internet Options > Connections to relocate your alternate DNS server (if any).
Refreshing the Connection Between Cisco Nexus 1000V and Microsoft SCVMM Server
You can refresh the connection between the Cisco Nexus 1000V and Microsoft SCVMM server.
Step 1 Launch Microsoft SCVMM UI.
Step 2 Choose Fabric Management > Networking > Switch Extension Manager.
Step 3 Choose Cisco Nexus 1000V and right click to refresh. See Figure 3-1.
Figure 3-1 Refresh Cisco Nexus 1000V Connection with Microsoft SCVMM Server
Step 4 Verify that the job is complete by checking the Jobs section.
Updating the Cisco Nexus 1000V Configuration Data on Hyper-V Hosts
You can update the Cisco Nexus 1000V configuration data on the Hyper-V hosts.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Navigate to Fabric and click Logical Switches to display the screen. See Figure 3-2.
Figure 3-2 Display Logical Switches
Step 3 Choose Hosts from the toolbar.
Step 4 Choose the host and choose 1000V. See Figure 3-3.
Figure 3-3 Select Nexus 1000V Switch
Step 5 Choose Remediate from the toolbar.
Step 6 Verify that the job was completed by checking the Jobs section.
Verifying Cisco Provider Installed Correctly
You can verify the Cisco Provider has been installed correctly.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Navigate to Settings.
Step 3 Choose Configuration Providers.
Cleaning Up Switch Extention Fails
This problem might occur when deploying a VM using a static IP address from the static IP pools published by the Cisco Nexus 1000V VSM.
Note This is a known Microsoft issue.
Since the error is due to unrevoked IP addresses, the error shown by the Microsoft SCVMM is not clear
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Using a PowerShell window, run the following commands, in sequence, to revoke the static IP adresses:
$vsem = Get-SCVirtualSwitchExtensionManager -VirtualSwitchExtensionManagerConnectionString
$pools = Get-SCStaticIPAddressPool | where { $_.VirtualSwitchExtensionManager.ID -eq $vsem.ID }
$pools | ForEach-Object { get-scipaddress -UnAssigned -StaticIPAddressPool $_ } | Revoke-
Refreshing Switch Extension Manager Fails
The following are symptoms, possible causes, and solutions for problems when refreshing the Switch Extension Manager.
|
|
|
Unable to refresh the Switch Extension Manager from Microsoft SCVMM. |
There is a problem with the connection between the Microsoft SCVMM and the VSM. |
1. Verify that you can navigate to the VSM http://<vsm_ip_address> from the server where the Microsoft SCVMM service is running. 2. Verify that your proxy settings and firewall settings are not impacting on Microsoft SCVMM to VSM connectivity. |
There is an error in the VSM configuration. |
On the VSM, verify the configuration by entering the show svs domain command. |
Verifying the Logical Switch Compliance
The Microsoft SCVMM might report a non-compliant warning when deploying or changing port profiles on the Cisco Nexus 1000V logical switch. This is a result of a mismatch of the opaque data stored on Microsoft SCVMM and that of the individual hosts.
Note This is only a warning—it is not an error.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Navigate to Fabric > Logical Switches > Hosts.
Step 3 Using a Microsoft SCVMM PowerShell window enter:
- Get-SCVirtualNetwork | where-object {$_.LogicalSwitch -like “1000V”} | select VMHost, HighlyAvailable, LogicalNetworks, VMHostNetworkAdaters | LogicalSwitchComplianceStatus
To remove the Logical Switch Compliance Warning:
Step 1 Refresh the Virtual Switch Extension Manager
Step 2 Navigate to Fabric > Logical Switches > Hosts.
Step 3 Select the appropriate Logical Switch and choose Remediate the Host.
Verifying the Logical Switch Extension
The Cisco Nexus 1000V logical switch extension is always a forwarding extension. You can verify the logical switch extension.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Navigate to Fabric > Logical Switches > switch_name > Properties > Extensions.
Step 3 Verify the extension type is Forwarding.
Verifying the Logical Switch Uplink Mode
The Cisco Nexus 1000V logical switch uplink mode should be team. You can verify the logical switch uplink mode.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Navigate to Fabric > Logical Switches > switch_name > Properties > Uplink.
Step 3 Verify the Uplink mode is Team.
Creating or Deleting a Switch on a Host Management Adapter
While deploying a Cisco Nexus 1000V switch or cleaning up a Cisco Nexus 1000V on a host management adapter, the operation may fail if there are network flaps or DNS resolution. This might cause host connectivity lose because the failure occurs on the host management adapter.
To recover the host management connectivity and clean up the stale switch:
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Log in to the host using the remote console.
Step 3 Using a Microsoft SCVMM PowerShell window, enter:
Remove-VMSwitch > name switchname
Step 4 Enter the following command to remove the NetSwitch Team from the host and restore connectivity:
Get-NetswitchTeam | Remove-NetSwitchTeam
Step 5 Refresh the host from the Microsoft SCVMM.
Note If Step 3 fails when the WMI on the host is stuck in an inconsistent state, manually delete the switch from the registry followed by a system reboot and proceed to Step 4.
Exporting VM Templates with a Hard Disk Fails
Exporting a VM template when you have hard disk selected fails. This is caused by internet proxy settings.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Verify that the internet Connection Settings field is blank.
Uninstalling a Microsoft SCVMM 2012 SP1 Fails
Uninstalling a Microsoft SCVMM 2012 SP1 fails.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Uninstall Update Rollout 2 (UR2) updates.
Step 3 Uninstall Microsoft SCVMM SP1 from Add/Remove Programs.
Deleting Temporary Templates
You can delete temporary templates that are created by Microsoft SCVMM.
|
|
|
Unable to delete Cisco Nexus 1000V objects in Microsoft SCVMM. |
Microsoft SCVMM creates temporary templates that are linked to the Cisco Nexus 1000V objects. |
1. Delete the temporary templates by entering the following commands in a PowerShell window.
- Get-VMMServer
- Get-SCVMTemplate | where {$_.Name -linke “Tempoarary*”} | Remove-SCVMTemplate
|
Verifying Host Compliance in Microsoft SCVMM
You can verify host compliance on a Microsoft SCVMM, all hosts should show as fully compliant.
Step 1 Launch the Microsoft SCVMM UI.
Navigate to Fabric > Logical Switches > Hosts.
Step 2 Select host from list.
Step 3 Choose Remediate from the toolbar.
Step 4 Verify that the job was completed by checking the Jobs section.
Creating a Switch on a Management NIC with Static IP Fails on a Server Core
Creating a switch fails when using a Cisco Nexus 1000V on a management NIC with a static IP address on a server core.
Note This is a Microsoft issue with Server Core versions of Windows Server 2012.
Step 1 Launch the Microsoft SCVMM UI.
Step 2 Login to the host using the remote console.
Step 3 Using a Microsoft SCVMM PowerShell window, delete the switch from the host by entering:
Remove-VMSwitch – name switchname
Step 4 Enter the following command to remove the NetSwitch Team from the host and restore connectivity:
Get-NetswitchTeam | Remove-NetSwitchTeam
Step 5 Refresh the host from the Microsoft SCVMM.
Problems with Management NICs
The following are symptoms, possible causes, and solutions for problems with management NICs.
|
|
|
Unable to push opaque data (OD) on VEMs. |
VSM IP address has changed. |
1. Change the IP address of the management interface (mgmt0) on the VSM. 2. Change the connection string of the Switch Extension on the Microsoft SCVMM to the new VSM IP address. 3. Refresh the Switch Extension Manager in Microsoft SCVMM. 4. Verify the information on all screens before selecting OK. 5. Navigate to Fabric > Logical Switches > Hosts. 6. Select the host from the list. 7. Choose Remediate from the toolbar. 8. Verify that the job was completed by checking the Jobs section |