I have summarized ha and dpm one after another. I have prepared tests over the past few days. drs has also studied the issue and I think it is time to summarize it for your reference. 1. drs concept VMwareInfrastructure 3 DRS (Distributed Resource schedted) Vmware Distributed Resource scheduling program (DRS) dynamically allocates and balances computing capacity across hardware resources aggregated to the logical Resource pool. Vmware DRS continuously monitors utilization across resource pools and intelligently allocates available resources among multiple virtual machines based on predefined rules that reflect business needs and changing priority transactions. When the load on a virtual machine increases, vmware DRS automatically allocates more resources between physical servers in the resource pool by re-distributing virtual machines. 2. Basic Steps for experiment Setup 1. Cluster-right-click-edit settings (we need to test drs effect, so set drs to manual)
2. Our tests can be divided into several types,
A: Do not set rules, B: Cluster virtual machine, C: independent virtual machine, D: Virtual Machine to the host.
A. Assume that we use three virtual machines as testing virtual machines, and run the loop script bat in the Virtual Machine. One or more of the three virtual machines will be migrated to other idle hosts.
B. Suppose we set these three VMS to a test group and run the loop script bat in the VM, 3 Virtual opportunities are migrated to a host with a large amount of idle resources (including the local machine. This method applies when there is a Dongle)
C. Suppose we set these three VMS to a test group and run the loop script bat in the virtual machine. The three virtual opportunities are migrated to other idle virtual machines.
D. Suppose we set the three VMS to a test group, create the three hosts as the pc host group, and run the loop script bat in the VM. Here there is a concentrated situation, of course, now and B, C set testing. And B set results: three virtual opportunities are migrated to one of the idle hosts in the pc host group.
3. In the above tests, you can see the results in the manual test. Because drs extracts data from the esxi host every 30 minutes, it is automatically balanced once every 8 hours. Therefore, we use manual testing. We recommend that you use automatic testing in the production environment. Author: yujin2010good