Cisco Video Surveillance Virtual Machine Deployment and Recovery Guide for UCS Platforms, Release 7

Logical Design

HA as part of vSphere 5.0, also referred to as FDM (Fault Domain Manager), introduces less complexity and higher resiliency than its previous releases. The FDM agent communicates with vCenter to retrieve information about the status of virtual machines and vCenter is used to display the protection status of virtual machines. HA relies on vCenter to be available to configure or monitor the cluster so if vCenter is unavailable, it will not be possible to make changes to the configuration of the cluster. By design, it will respond to failures without vCenter, thus not affecting the availability of the protected virtual machines.

Figure 3-3 describes the communication in a vSphere HA environment.

Figure 3-3 VMware HA Design

When HA is enabled, the HA agents on all hosts participate in an election to pick a master. The remaining hosts act as slaves. The master communicates with the vCenter server to report state information of the cluster. The slaves talk to the master server to forward monitoring information of VMs running locally on that host.

vSphere HA protects VMs against the following failures:

• ESXi host failure
• Virtual Machine failure

Checklist

To implement vSphere HA, the following requirements should be met:

• All hosts in a vSphere HA-enabled cluster must have access to the same shared storage locations used by the VM on the cluster. This includes any Fibre Channel, FCoE, iSCSI, and NFS datastores used by the VM.
• All hosts in a vSphere HA cluster should have an identical virtual networking configuration. If you are using a vSphere Distributed Switch (vDS), all hosts should be participating in the same vDS.
• An easy and simple test for identifying HA capability for a virtual machine is to perform a vMotion. The requirements of vMotion are actually more stringent than those for performing an HA failover, though some of the requirements are identical.

Cluster Configuration

vSphere HA is a cluster-level configuration. To use vSphere HA to protect the VSM VM, you must first place your ESXi hosts into a cluster.

Perform the following procedure to the ESXi hosts into a cluster:

Procedure

Step 1 Login to the vCenter server using the vSphere client.

Step 2 Navigate to Home > Inventory > Hosts and Clusters.

Note You must have administrative privileges to configure a cluster and vSphere HA.

Step 3 Right-click on the Datacenter level of the Inventory hierarchy and choose New Cluster (see Figure 3-4).

Figure 3-4 Creating a New Cluster

Step 4 Enter a cluster name.

Step 5 Check the Turn on vSphere HA check box to enable HA on this cluster.

Step 6 Click Next on subsequent screens to accept default settings and finish.

Figure 3-5 Creating a New Cluster

Step 7 Drag and drop the ESXi hosts that are required to participate in vSphere HA onto this cluster. Be sure to include the host running the VSM VM into this cluster.

Configuring vSphere HA Parameters

To configure vSphere HA cluster parameters, right-click on the cluster and choose Edit Settings, click vSphere HA on the left, and proceed to the following sections:

• Host Monitoring
• Admission Control
• Virtual Machine Options
• VM Monitoring
• Datastore Heartbeating

Host Monitoring

In the Host Monitoring Status area, check the Enable Host Monitoring check box (see Figure 3-6). This enables the cluster to monitor for ESXi host failures and initiate a restart of the VSM VM upon a host failure.

Figure 3-6 vSphere HA Parameters

Admission Control

Admission Control determines the amount of the cluster resources that is reserved for failovers. Reserving higher capacity will allow for tolerating multiple failovers but it also reduces the amount of VMs that can be run.

This can be enforced by selecting Enable in the Admission Control area (see Figure 3-6). As a result, if sufficient resources are not available at the time of a failover, the VM will not be allowed to power on. This avoids over-allocation of resources to VMs.

Cluster resources to be reserved can be configured in the Admission Control Policy area (see Figure 3-6). They can be reserved in terms of:

• Host failures —The number of ESXi host failures that the cluster can tolerate.
• Percentage of cluster resources —Percentage of CPU and Memory resources within the cluster that will be reserved to power on VMs in case of a failover.
• Specify failover hosts —Manually specify which hosts will be reserved totally for restarting VMs.

Note The above policy settings depend on cluster capacity, number of VM running, and resource reservation (at a VM level). If the cluster is running only the VSM VM, then the following configuration is recommended:

• Choose Percentage of cluster resources reserved as Failover space capacity in the Admission Control Policy area.
• Set 25% CPU resources
• Set 25% Memory resources

Virtual Machine Options

Virtual Machine Options (see Figure 3-7) can be configured at the cluster level and the virtual machine level. Configuration at the virtual machine level overrides cluster-level settings. The following settings can be configured:

• VM Restart Priority —Priority to be used for VMs during a failover.
• Host Isolation Response —The isolation response refers to the action that vSphere HA takes when the heartbeat network is isolated.

To configure the settings that correspond to the VSM VM, choose the following settings in the ‘Virtual Machine Settings area:

• VM Restart Priority— High
• Host Isolation Response:

– Shut Down —Select this option when using network based storage protocols like iSCSI, NFS.

– Leave Powered On —Select this option for other environments.

Figure 3-7 Virtual Machine Options

VM Monitoring

VM Monitoring determines whether the VM is monitored for its VMware Tools heartbeats or VMware Tools application heartbeats.

In the VM Monitoring Status area, choose VM Monitoring only from the drop-down menu in the VM Monitoring Status area.

Three default levels of sensitivity are provided – Low, Medium and High – along with a Custom level to manually set the sensitivity values. The below table explains these levels.

To set Monitoring Sensitivity settings corresponding the VSM VM, choose the appropriate option from the VM Monitoring drop-down menu in the Virtual Machine Settings area. See Figure 3-8 and Table 3-1 for details.

Figure 3-8 VM Monitoring

Datastore Heartbeating

In case of management network failure, vSphere HA monitors hosts and VMs using heartbeats to datastores. A minimum of 2 datastores are required for this purpose.

Navigate to the Datastore Heartbeating page (see Figure 3-9) and choose Select any of the cluster datastores taking into account my preference, and choose the preferred datastores from the list provided below. Click OK to save settings.

Figure 3-9 Datastore Heartbeating

Cluster Status

Cluster-level status of HA can be monitored by clicking on the appropriate cluster in the inventory hierarchy and navigating to the Summary tab (see Figure 3-10). Under the General section the status of vSphere HA, total resources that are part of the cluster (CPU, memory, storage), total vSphere entities in the cluster (hosts, processors, datastores, VMs, templates) are displayed. Figure 10 below shows the Summary for a sample cluster.

More importantly cluster parameters are displayed under the vSphere HA section. Click on the Cluster Status link to check the status of the cluster. Details such as Cluster master host, number of slave hosts, and number of protected VMs can be viewed from this link.

Figure 3-10 VMware HA Cluster Status

Host Status

Status of ESXi hosts participating in the HA cluster can be monitored by clicking on the appropriate host in the inventory hierarchy and navigating to the Summary tab. Under the General section, vSphere HA State shows the host status. Figure 3-11 shows the status of a sample host in a cluster. Important states to note are:

• Running (Master)—This host is active and the master host in the cluster.
• Connected (Slave)—This host is active and one of the slave host(s) in the cluster.

Figure 3-11 VMware HA Host Status

Virtual Machine Status

The status of individual VMs can be monitored by clicking on the appropriate VM in the inventory hierarchy and navigating to the Summary tab. Under the General section, vSphere HA Protection shows the VM protection status. Figure 3-12 shows the status of a sample VM in a cluster.

Figure 3-12 Figure 12. VMware HA VM Status