As I know VMware HA is focused on hardware failure,not on operating system or software failure.(mentioned in vmware_ha_wp.pdf)
At the same time I found following feature as mentioned in VMware-High-Availability-DS-EN.pdf
Automatic detection of operating system failures ->>
VMware HA detects operating system failures within virtual machines by monitoring heartbeat information. If a failure is detected, the affected virtual machine is automatically restarted on the server.
Above two stmt are contradictory.Can you please explain it.
By default HA does not monitor individual VMs - it is a feature that can be enabled. When enabled it will look for the heartbeat that VMware Tools sends and if the heartbeat is not detected the VM will restart - you can set the sensitibity to which HA will respond to the loss of heartbeat so the VM does not restart just beacuse the machine is overloaded -
Hi,
Like weinstein5 says, HA can detect when guest OS fails and can restart VM automatically, but be aware that this feature us related only to OS not for APP, this feature it isn't APP aware - just for your information. I saw many discussion on forums about that feature and a lot of confusion around it. There are some third party tools which leverage that HA feature and are APP aware (it was something from Symantec - forgot a name of the product)
That feature relay on VMware tool heartbeat but not only, also on CPU activity, LAN and SAN traffic, if they are below certain threshold then HA consider VM as a faulty and try to restart it.
Thanks guys.Till I have confusion.What type of OS failure can it detect?
Say,my OS crashed...natuarally it won't get heartbeat from that VMs(am I right?).
Then will HA comes in picture?
Can you please give some example.
Hello there,
When your OS crashes, your VMware tools will not be responding. This is when your HA will intiate a reboot.
Regards,
rsaha wrote:
Thanks guys.Till I have confusion.What type of OS failure can it detect?
Say,my OS crashed...natuarally it won't get heartbeat from that VMs(am I right?).
hi
When your OS creashed, VMtools stops responding to HA, and also CPU, LAN and DISK activity will indicate that OS crashed
Thanks guys.Till I have confusion.What type of OS failure can it detect?
All OS versions supported by VMware as guest OS, one requirments is to have VMtools installed on guest OS
If my OS crashed.Then what is the use of HA?.I need to rebuild the guest OS.
So how HA works?
A OS crashed does not mean that the OS must be reinstalled... Could also be an exeption (for example).
Andre
Arturka, the Symantec product you are talking about is ApplicationHA.
App monitoring is actually a new feature of vSphere 5.
Here is a post from yellowbricks
http://www.yellow-bricks.com/2011/08/11/vsphere-5-0-ha-application-monitoring-intro/
Hai,
HA is not a fault tolerant(provides uninterruptable access). It is a reactive system, which react to a problem(failed host). So it doesnot repair the faied host it does what required after a failure occures to host and lead to restarting VM's on other host which is having plenty of resources to hold VM's.
At the same time I found following feature as mentioned in VMware-High-Availability-DS-EN.pdf
Automatic detection of operating system failures ->>
__________________________________________________
Agreed,but unable to understand above stmt in VMware-High-Availability-DS-EN.pdf.
as mentioned earlier.
What is meant by "Operating System Failure" is usually a blue screen in Windows or a panic in Linux. HA monitors the VMware Tools heartbeat as well as network and disk activity of the guest. If neither are detected, HA powers off the VM (like pulling the power plug) and tries to restart the VM.
André