Virtual Server Frequently Asked Questions (FAQs)
What is included in the base virtual server cost?
The base virtual server fee
includes:
-
The use of VMware® Virtual Infrastructure 3 (ESX and Virtual Center)
-
One virtual central processing unit (CPU)
-
The first gigabyte of memory
-
One copy of Windows Server (Standard, Enterprise, 32-bit, or 64-bit)
-
Access to the server console through the VMware Virtual Center
-
The use of VMware's VMotion to move servers from one physical host to another
without disruption
-
Load balancing to ensure virtual server workloads are evenly distributed across
physical server hosts. Balanced workloads function dynamically and without
disruption
-
The ability to restart servers on an alternate physical server host if a server
fails or needs to be taken out of service for maintenance
-
DIS maintenance and monitoring of the virtual server environment
-
24x7 technical support
The following services can be provided
at an additional charge:
-
Storage is available in four classes: EMC2 Symmetrix or EMC2 Clarion in either
RAID 1 or RAID 5; storage can be allocated from one or more classes depending
on requirements (e.g. RAID1 for a database but RAID5 for the “C drive”)
-
Server backup/recovery services (same as for physical servers)
-
Server monitoring (same as physical servers)
-
Additional virtual CPUs and memory
-
Technical support in the Washington State Data Center
What is a virtual central processing unit (CPU)?
Each virtual machine or server is assigned one, two, or four virtual CPUs. More
than 80% of all virtual servers do not need more than one virtual CPU. DIS
Virtual Server Hosting uses host servers with Intel quad core processors. The
VMware hypervisor can assign these processor cores independently to any active
virtual server. Each processor core is acting as an independent CPU. When a
virtual server is defined with one virtual CPU, the virtual server can use up
to one processor core at a time. If the server is defined with two virtual
CPUs, it uses two processor cores when it is active. With four virtual CPUs,
the virtual server uses four processor cores when it is active.
What is the advantage of VMware Clusters?
DIS uses Virtual Infrastructure 3 software in a clustered configuration. This
configuration provides several advantages:
-
Virtual servers may be moved from one hardware platform to another
non-disruptively
-
Virtual servers will be load-balanced across multiple hardware platforms
depending on the workload of each hardware platform
-
When maintenance is scheduled for a hardware platform, the server workload can
be moved off of the affected platform in advance to eliminate service outages
-
If there is a need to move a virtual server to a new storage device, it can be
done non-disruptively
-
If there is a need to move a server to a new hardware platform during a server
upgrade, there is no disruption
-
If a hardware platform fails, the virtual servers on that platform will be
automatically rebooted on another hardware platform
How is memory defined and billed?
A virtual server is allocated with a set amount of memory in increments of one
gigabyte. The amount of memory actually being used is regulated by the VMware
hypervisor, which can consolidate unused memory, duplicate memory pages and
make them available to other virtual servers. As the amount of memory used can
vary from day to day, DIS will assess fees according to how much memory is
defined rather than how much memory is actually used.
What kinds of storage are available?
Virtual server disk storage costs are priced and billed separately because
there are several different classes of storage to choose from.
DIS offers two tiers of storage:
-
The fastest and most fault-tolerant is based on the EMC2® Symmetrix DMX
equipment, which can withstand the failure of any component without any
perceived loss of performance.
-
The mid-tier class of storage is based on the EMC Clarion CX equipment, which
provides high performance but has less redundancy. It can survive the failure
of any component, but if certain components fail, there will be degradation in
performance.
Each tier of storage can be implemented as RAID 1 (mirrored) which is fault
tolerant and has the highest performance or RAID 5 which is also fault tolerant
with a lower level of performance. All storage used in the virtual server
hosting service offering is SAN storage.
This allows storage to be shared between physical hosts so that virtual server
movement (VMotion), load balancing, and restart services can be deployed.
What if a server becomes too busy?
DIS Virtual Server Hosting through the VMware hypervisor gives each server a
“share” of the workload. If one server becomes very busy, it will use all of
the processor capacity it can. If two servers become very busy, they will each
get half of the resources. If a server has a pattern of being very busy, either
more resources in memory and virtual CPUs will need to be defined, or the
server should be converted to a physical server.