Advanced Configuration options for VMware High Availability
Purpose
In the majority of environments, VMware High Availability (HA) default settings do not need to be changed. Depending on your specific environment you may need to modify some HA options. The sections below describe the different configuration options available and how to apply them.
This article explains Advanced Configuration options for VMware High Availability (HA).
Resolution
Note: Not all configuration variables work in all versions of VirtualCenter. As new variables are introduced in newer releases, they remain throughout later versions.
Applying a VMware HA customization
To access the various customizations available:
Login to VirtualCenter with the VI Client as an administrator.
Right click the Cluster in the inventory.
Select VMware HA as shown below:
Click the Advanced Options button and the following window will appear:
Enter Option and Value fields as appropriate (see below).
Click OK.
Click OK again.
Wait for the Reconfigure Cluster task to complete and then right click the Cluster again from the inventory.
Click properties.
Disable VMware HA and wait for the Reconfiguration Cluster task(s) to complete.
Right click the cluster and Enable VMware HA to have the settings take effect.
VMware High Availability Advanced options
The following are settings available in VirtualCenter 2.0.0 and above:
das.defaultfailoverhost = <hostname> - Add this option/value pair is used to set a preferred failover host, where <hostname> represents the short name of the preferred host. The host specified will be used as the first choise incase of a failure, however if there is insufficient capacity VMware HA will automatically select an alternate host.
das.isolationaddress = <value> - Add this option/value pair change the default isolation address, where <value> represents the IP address that is to be used as the isolation address. By default the address used is the Default Gateway address for the ESX Server because it is a highly available address. This address is queried as one of the items in determining if an isolation response has occurred on the host, thus if this parameter is changed be sure to use a reliable, highly available IP address with as few failure points between it and the ESX Server as possible.
The following are settings available in VirtualCenter 2.0.2 and above:
das.failuredetectiontime = <value> - Add this option/value pair to change the default failure detection timeout, where <value> represents the failure time in milliseconds. VMware HA uses this timeout in declaring an isolation response, and will not declare a host as isolated until this timeout has been reached without any heartbeats recieved. By default the default failure detection time is 15 seconds (15000 ms), however another common alternative is 60 seconds (60000 ms).
das.usedefaultisolationaddress = <value> - add this option/value pair to disable the use of the default gateway as an isolation address, where value is either true or false. By default this value is set to true. This parameter is generally used in conjunction with the das.isolationaddress1 to das.isolationaddress10 parameter(s) listed below.
das.isolationaddress1 to das.isolationaddress10 = <value> - add these option/value pair(s) to specify more than one alternate isolation address for VMware HA to use, where <value> represents the IP address to be used for isolation detection. When using more than one isolation address it is recommended that the das.failuredetectiontime parameter be incresed to ensure proper failover detection can occur. Also, although up to 10 different isolation addresses can be specified one or two addresses should be sufficient for proper failover detection. For more information on setting multiple isolation addresses, see Setting Multiple Isolation Response Addresses for VMware High Availability (1002117).
Note: the das.isolationaddress parameter is still supported, if you are not using more than one alternative isolation address.
The following are settings available in VirtualCenter 2.5.0 and above:
das.powerOffOnIsolation = <value> - add this option/value pair to overide the default isolation response setting to shutdown all the virtual machines when an isolation response occurs, where <value> represents either true or false. By default this value is set to true. If changed to false all virtual machines will remain powered on in case of isolation.
das.vmMemoryMinMB = <value> - add this option/value pair to overide the default memory slot size value used for admission control for VMware HA, where <value> is the amount of RAM in MB to be used for the calculation if there are no larger memory reservations. By default this value is set to 256MB.
das.vmCpuMinMHz = <value> - add this option/value pair to overide the default CPU slot size value used for admission control for VMware HA, where <value> is the amount of CPU in MHz to be used for the calculation if there are no larger CPU reservations. By default this value is set to 256MHz.
The following are settings available in VirtualCenter 2.5.0 Update 2 and above:
das.conservativeCpuSlot = <value> - add this option/value pair to revert how conservative VMware HA is in regards to vCPU when calculating slot size for admission control, where <value> represents either a value of true or false. By default in VirtualCenter 2.5.0 Update 2 this option is false to make the admission control less conservative, if set to true this will revert to the behavior of admission control as it was VirtualCenter 2.5.0, and VirtualCenter 2.5.0 Update 1.
das.allowVmotionNetworks = <value> - add this option/value pair to allow VMotion networks to be considered for VMware HA usage, where <value> represents a value of true or false. This parameter will enable a host that only has one Network card configured for management and VMotion combined to be properly configured for VMware HA. By default this value is false, and any VMotion network is ignored.
das.allowNetwork[...] - is another advanced option that enables the use of port group names to control the networks used for HA. You can set the value to be "Service Console 2" or "Management Network" to use those port group names in the networking configuration.
