Introduction to the HP Virtual Server Environment

Conveniently expand and contract application server pools: Suppose a company has an
application consisting of several servers or nodes in a cluster configuration whose load varies
over the course of a business cycle. By using logical servers, the operations group can “dial-
up or down” more or less blade servers to meet the changing demand.
Re-use server resources for another purpose: A company may have multiple Q/A or test
systems for different applications. If they are not all needed at the same time, logical servers
can be associated with each one of them. When one is no longer needed, it can be
deactivated, which will free up the blade for use by another logical server. All of the specific
configuration information associated with this logical server is maintained in the logical server
profile, including it’s MAC address and WWNs. So, it can be re-activated at a later time.
Another logical server can now be activated to re-use that server blade.
Move the Logical Server to a different location for maintenance purposes: Another case is
that a blade needs to be taken down to upgrade the firmware or something like that. Just
move the Logical Server to a different blade, perhaps even one in a different enclosure as
long as it’s in the same Virtual Connect Domain. (Of course, the operating system does have
to shut down and be rebooted).
Move the Logical Server for load balancing purposes: VM logical servers can easily be
moved from one hypervisor to another in order to balance the load between them. The
Capacity Advisor software has a load balancing feature that can make recommendations for
the best way to place the VMs. There is even a drag and drop capability within
Virtualization Manager which can be used to move them, and the 5-star rating system will
show you the best target locations.
HP Global Workload Manager
Global Workload Manager, described briefly in an earlier section, is a part of the HP Virtual Server
Environment software suite. The gWLM software is designed to manage a large number of servers
and workloads in a virtualized environment. To accomplish this, gWLM runs on a CMS and is
integrated with HP SIM and Virtualization Manager. From the CMS, you can centrally manage the
servers, partitions, and workloads that gWLM manages. You can apply policies to workloads that
govern both when and how resources are shared. Also, you can get reports, both real-time and
historical, about how the resources were allocated, how many resources each workload used, and
how many each workload needed. The gWLM software operates in “advisory mode” as well as in
“managed mode.” In advisory mode, rather than automatically move CPU resources, gWLM informs
you that it would have done so. One of the design goals for gWLM was to allow users who are
already using Virtual Partitions to understand, install, and configure gWLM to manage those Virtual
Partitions in less than 30 minutes.
24