Caution: Increasing the VMware HA timeout value also increases the amount of time before real failures are detected. A longer failure detection time may cause a longer outage within your environment. Changing the HA timeout also affects the isolation detection time and the small gap during which the host regains network connectivity before the isolation detection time is activated. Perform these steps at your own risk.
The method for adjusting the timeout depends on which version of VirtualCenter you are using to manage your ESX host. It is important to use the correct instructions for your version of VirtualCenter. Express the new timeout value in milliseconds. To convert seconds to milliseconds, multiply by 1000. For example, the default timeout value 15 seconds = 15000 milliseconds.
VirtualCenter 2.0.x
To increase the VMware HA timeout value for VirtualCenter 2.0.x:
- Log in to the ESX host console command-line as root.
- Run the following command to see the timeout value:
/opt/LGTOaam512/bin/ftcli -cmd getFailureDetectionInfo
- Run the following commands to increase the timeout value:
Note: FT_DOMAIN is the VMware HA cluster domain. It has nothing to do with your network domain. Enter the value as shown.
FT_DOMAIN=vmware
FT_DIR=/opt/LGTOaam512
export FT_DOMAIN
export FT_DIR
/opt/LGTOaam512/bin/ftcli -cmd "setFailureDetectionTime XXXXX"
Note: Replace XXXXX with the new timeout value in milliseconds.
VirtualCenter 2.5.x and vSphere 4.0.x
To increase the VMware HA timeout value for VirtualCenter 2.5.x and vSphere 4.0.x:
-
In VMware Infrastructure Client (VI Client), click on the HA cluster.
-
In the Summary tab, click Edit Settings.
-
In the Settings dialog box, deselect Enable VMware HA and click OK.
Note: You must wait for all reconfigure and unconfigure tasks to finish.
-
Click on the HA Cluster.
-
In the Summary tab, click Edit Settings.
-
Select Enable VMware HA.
-
In the left pane, click VMware HA.
-
Click Advanced Options.
-
In the Advanced Options (HA) dialog box:
-
In the option name field, enter das.failuredetectiontime .
-
For the value, enter the new timeout value in milliseconds.
Note: The default value for das.failuredetectiontime is 15000 milliseconds (15 seconds).
-
Click OK.
-
Click OK.
Note: You must wait for all reconfigure and configuring tasks to complete. The change propagates to the ESX hosts in the cluster during HA configuration.