虽然没试过HA下VCSERVER是不是正常动作,但从我目前看到的材料上来看是应该能正常工作的,理由有两个:
1.HA不依赖VCSERVER就能正常工作;
2.VCP510考题中就有一题关于VCSERVER虚拟机在HA下能正常重启的考题,说明这个功能能实现。
QUESTION 17
vCenter Server is running in a virtual machine that is part of a vSphere HA and SDRS cluster. The vCenter Server virual machine has been tested an can successfully migrate with vMoaon between all ESXi hosts m the cluster DRS is configured for partial automation.
What happens if an ESXi host that is running the vCenter Server virtual machine experiences a power outage?
A. All DRS cluster operations will be unavailable until the failed ESXi host is brought back online.
B. vCenter Server will remain offline, but all virtual machines on the remaining hosts will continue without interruption.
C. This situation cannot occur, because vCenter Server cannot be installed on a virtual machine in a vSphere HA cluster.
D. vSphere HA will restart the vCenter Server virtual machine on another host
Answer: D
刚才打开了高手的5.0 CLUSTER深度研究一书里面就有VMOTION的介绍
说得很生动
What does the workflow of a vMotion process look like?
1. Validate destination host is compatible.
2. Create a shadow virtual machine on the destination host.
3. Copy each memory page from the source to the destination via the vMotion network. This process is known as pre-copy.
4. Perform another pass over the virtual machine’s memory, copying any pages that changed during the last pre-copy iteration.
5. Continue this iterative memory copying until no changed pages (outstanding to be-copied pages) remain.
6. Stun the VM on the source and resume it on the destination.
尤其第6项,STUN(打晕)原VM,然后恢复目标机器那边那个VM 所以这个过程也可以理解成中断了(尤其是那些IO高的VM,举例高速的数据库或者视频点播服务器)切换的过程一定会断的
在5.0之后的VMotion
Changes to vMotion in vSphere 5.0?
There are some fundamental changes when it comes to vMotion scalability and performance in vSphere 5.0 which we have listed below:
Multi-NIC vMotion support
Stun During Page Send (SDPS)
Higher Latency Link Support
Improved error reporting