Information Technology Reference
In-Depth Information
Storage vMotion
Storage vMotion is the storage equivalent of vMotion, and it is used to
manually balance storage utilization between two datastores.
vSphere Distributed Resource Scheduler
vSphere Distributed Resource Scheduler (DRS)
is used to automatically balance compute resource utilization among two or more ESXi hosts.
Storage DRS
Just as Storage vMotion is the storage equivalent of vMotion, Storage DRS
is the storage equivalent of DRS, and it is used to automatically balance storage utilization
among two or more datastores before and after initial placement of virtual machine i les.
As we explain each of these four mechanisms for balancing resource utilization, we'll also
introduce or review a few related features of vSphere, such as clusters and VMware Enhanced
vMotion Compatibility (EVC).
Let's start with vMotion.
Exploring vMotion
We've dei ned the vMotion feature as a way to manually balance compute resource utilization
between two ESXi hosts. What does that mean, exactly? vMotion can perform a live migra-
tion of a VM from one ESXi host to another ESXi host without service interruption. This is a
no-downtime operation; network connections are not dropped and applications continue run-
ning uninterrupted. In fact, the end users are unaware that the VM has been migrated between
physical ESXi hosts. When you use vMotion to migrate a VM from one ESXi host to another, you
also migrate the resource allocation—CPU and memory—from one host to another. This makes
vMotion an extremely effective tool for manually load-balancing VMs across ESXi hosts and
eliminating “hot spots”
—
heavily utilized ESXi hosts—within your virtualized datacenter.
In addition to manually balancing VM loads among ESXi hosts, vMotion brings other ben-
ei ts. If an ESXi host needs to be powered off for hardware maintenance or some other function
that would take it out of production, you can use vMotion to migrate all active VMs from the
host going ofl ine to another host without waiting for a hardware maintenance window. Because
vMotion is a live migration—no interruption in service and no downtime—the VMs will remain
available to the users who need them.
While it sounds like magic, the basic premise of vMotion is relatively straightforward. vMo-
tion works by copying the contents of VMmemory from one ESXi host to another and then
transferring control of the VM's disk i les to the target host.
Let's take a closer look. vMotion operates in the following sequence:
1.
An administrator initiates a migration of a running VM (VM1) from one ESXi host (esxi-
03a) to another (esxi-04a), as shown in Figure 12.1.
2.
The source host (esxi-03a) copies the active memory pages VM1 has in host memory to
the destination host (esxi-04a) across a VMkernel interface enabled for vMotion. This is
called
preCopy
. Meanwhile, the VM still services clients on the source (esxi-04a). As the
memory is copied from the source host to the target, pages in memory can be changed.
ESXi handles this by keeping a log of changes that occur in the memory of the VM on
the source host after that memory address is copied to the target host. This log is called a
memory
bitmap
. See Figure 12.2. Note that this process occurs iteratively, repeatedly copy-
ing over memory contents that have changed.
