Trends in

Server Virtualization

introduction
Server virtualization has come a long way in a recovery. Failure of pilot project apps wouldnt
very short time. From its early days in IT test and have a financial impact on the business and
development pilot projects, with VMwares vSphere administrators would have time to figure out how to
being the only game in town, to mass adoption, recover apps and data.
virtualization-first IT policies, and a range of
Jump forward a few years and its now routine
hypervisors available from companies like Citrix,
to move mission-critical applications into virtual
Microsoft, and Red Hat Linux.
machines. Applications like Microsoft Exchange,
Server virtualization has now found its way into IT Microsoft Sharepoint, Microsoft SQL Server, Oracle,
infrastructures of just about every size. A recent and SAP, invariably have demanding availability and
survey of IT professionals by Spiceworks found 80% data integrity needs. But, often the tools being used
of small- and mid-sized businesses have already for backup and recovery were not built for the new
adopted some form of server virtualization. and unique requirements of virtualized servers.

Virtualization is now viewed as a mainstream To complicate matters, applications running in

architecture, with many companies committed to virtual machines now frequently share resources.
deploying all new apps in virtualized environments. This builds dependencies into the architecture that
But it wasnt always that way. The earliest apps are often not given much attention when planning
running on virtualized servers tended to be less backup tasks. All of this requires rethinking the
critical to the business. Web and file servers were virtual server data protection strategy to make sure
common candidates. recoverability is aligned with the needs of individual
applications and the overall business goals of the
Pilot environments have less demanding backup
organization.
and recovery needs and administrators often relied
on existing backup and recovery tools, without
considering how virtual environments would affect

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 1
table of contents

CHAPTER 1 Trends in Data Center Virtualization

CHAPTER 2 Why Virtualize?
2.1 Reduced Capital Expenditure
2.2 Reduced Operating Expenses

CHAPTER 3 Virtualizations Downside

3.1 Not Everything Can Be Virtualized
3.2 Increased Cost
3.3 Server Sprawl
3.4 Single Point of Failure

CHAPTER 4 Backup Issues Unique to Virtualization

4.1 Backup Issues For Small & Mid-Sized Companies
4.2 Agentless Backup

CHAPTER 5 New Backup & Recovery Capabilities

5.1 In-Place Recovery
5.2 VM Migration
5.3 Changed Block Tracking
5.4 Advanced Capabilities

CHAPTER 6 The Business of Data Recovery

6.1 Recovery Point Objective (RPO)
6.2 Recovery Time Objective (RTO)

CHAPTER 7 Virtualization, Data Protection and Business Continuity

7.1 Its About The Business
7.2 Downtime Costs
7.3 Measuring Risk
7.4 Workable Plans
7.5 Fallback

CHAPTER 8 Virtualization and Tape

8.1 Industry Standards
8.2 Cost and Durability
8.3 Portability
8.4 Good Enough for Google

CHAPTER 9 Data Protection and the Hybrid Data Center

9.1 When is HA Not HA?

CHAPTER 10 Arcserve Unified Data Protection

10.1 Arcserve UDP and Magnetic Tape
10.2 Arcserve UDP 7000 Appliance Series

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 2
1

Trends in
Data Center
Virtualization
To read the technology press, you could be forgiven for
thinking that virtualization is now the entirety of the IT
computing landscape. This simply isnt true.
Sure, a lot of companies have now adopted The inevitable fallout from flat IT spending and
virtualization-first policies that dictate all new apps reduced IT staffing is that projects involving
must run in virtual environments. Moving legacy legacy applications get lower priority. Yes, newly
applications to virtualized environments, however, developed apps get to live on virtual servers,
is another matter entirely. courtesy of the dictates of virtualization-first
policies. But, the mantra of if-it-isnt-broke-dont-
Despite an increasing focus on technology fix-it will ensure that, short of natural end-of-life
and the benefits of IT to an organizations (EOL) driven spending, there is often no budget to
bottom-line, a recent survey by Spiceworks move legacy applications on physical servers into
found that IT budgets are at a the virtual world.
standstill. IT hiring is also not
keeping track with demand Most organizations, regardless of size, are now
for new technology. As a dealing with hybrid IT environments containing
consequence of these both virtual and physical servers. This is not an
trends, IT staffs ideal situation, by any means. Hybrid environments
are continually complicate just about every aspect of server
expected to do administration. They reduce administrator
more with less. productivity, increase the cost of management
tools, spread the knowledge of management tools
too thinly among administrators, and can lead
to compromised application availability and data
integrity.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 3
2

Why Virtualize?
So, what is all the fuss about virtualization?
Why are IT organizations so enamored
with this technology? The simplest answer
is that it saves money.
When business applications are deployed on The capacity estimate has to anticipate future
physical servers, administrators and capacity growth in demand for the application and peaks in
planners put their heads together to figure out demand. The need to provide adequate capacity for
how big a server is needed for the application. growth means that physical servers inevitably run
They make estimates for how much processing at, on average, 20 percent of their rated capacity.
power, how much memory, how much This means that 80 percent of the capital invested
storage, and how much network bandwidth the in server processor power, memory, storage, and
application needs. network capacity sits idle.

2.1 Reduced Capital Expenditure

Virtualization enables a physical server to host more physical server, each with their own discreet operating
than one virtual server. It also provides the capability to environment, can dramatically reduce the number
easily move virtual servers between different physical of servers in the data center. With fewer physical
servers to balance demand for resources. The physical servers, IT organizations are able to reduce their capital
servers running virtualization software frequently expenditure, freeing these funds for use elsewhere in
run at above 80 percent of their rated capacity. The the organization to increase revenue growth. Of course,
consolidation of business applications on a single there are other benefits, too.

2.2 Reduced Operating Expenses

Reducing the number of physical servers in the data do if setting up a physical server. This often reduces
center also saves energy, an important consideration provisioning from weeks to hours, or less, benefiting
when carbon-footprint is a metric tracked by investors rapid application development.
and shareholders. And, it enables a data center to host
more applications, a critical factor when data center Virtual machines are also simpler to administer than
real estate is becoming more valuable. their physical counterparts. A single virtualization
administrator is able to manage many more machines
From an administrative perspective, virtual machines than if they were looking after physical devices. This
are much easier to set up, and break down. If an benefits administrator productivity and can help alleviate
application needs a new server, an administrator can staffing issues.
provision the resources much faster than they could

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 4
3

Virtualizations
Downside
There are, of course, downsides to virtualization,
too. Not all business applications are appropriate for
running on a virtual server. While there are cost savings
to virtualization, the technology can also lead to increased
expenditure. As with many technologies, uninformed use can
exasperate the very problems that it is intended to solve.

3.1 Not Everything Can Be Virtualized

Not all applications are great candidates for virtualization. to the majority of application use cases, these unusual
Applications that are very sensitive to performance may applications are often not supported.
not be a good fit. These apps are unlikely to tolerate
sharing physical resources with others and the overhead Not all application software is capable of being virtualized.
of running a hypervisor on the same hardware may be For some, it can be licensing agreements limitations that
unwelcome. prevent virtualization. For others, it could be complexity.
Many organizations have older legacy applications that are
There is a wide variety of applications that require physical mission-critical, but are so complex due to many years
appendages to their servers, often with unique driver of upgrades and changes that moving them to a virtual
software. Because hypervisor software has to appeal platform would be too risky.

3.2 Increased Cost

There is a cost component that can impact the adoption performance needs of all virtual servers. These servers
of virtualization. While virtualization can reduce operating are likely to be more costly than the physical servers they
costs in the long-term, there are replace.
up-front expenses associated with
implementing the technology. Server and network administrators must be trained in the
art of virtualization. A wide variety of tools are available;
The host servers used to run each many provided by the hypervisor vendor.
virtualization hypervisor must
be capable of supporting the

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 5
3
3.3 Server Sprawl
Ironically, server sprawl a condition that virtualization Server virtualization addressed this problem. Consolidating
holds out the promise of solving can, in fact, be many physical servers on a single virtual server alleviates
exasperated by the ease of spinning up virtual machines. energy and floor-space constraints. However, the ease of
Server sprawl became a significant issue in the data center provisioning virtual machines can lead to renewed server
when servers were being deployed without a sufficient` sprawl.
understanding of their impact. This often resulted in
data centers full of under-utilized server hardware that
consumed precious energy and floor-space.

3.4 Single Point of Failure

Finally, a glaringly obvious downside to server Ensuring data availability and data integrity in a virtual
virtualization is the fact that hosting multiple virtual server environment demands a new approach to data
servers on one piece of hardware introduces the protection. Although many virtualization deployments rely
potential for a single point of failure. If the physical server on existing data protection techniques, assuming what
running the hypervisor fails, all applications running on worked for physical servers will work for virtual servers,
virtual machines hosted by the hypervisor will become virtualized infrastructures pose challenges. Not least of
unavailable. these is that the environment is unlikely to be all virtual.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 6
4

Backup Issues
Unique to
Virtualization
As applications One of the goals of server
like ERP, CRM, and virtualization has been to
email have moved make more efficient use
to virtual machines, of physical server hardware.
data protection Previous architectures would
in the virtualized use perhaps ten to thirty percent of
environment has a server's CPU, on average, leaving plenty
undoubtedly become of capacity for periodic workloads like backup.
more important. Unlike less critical applications, Virtualized server environments now typically see
these new workloads often have no tolerance utilization greater than 80 percent. This leaves very
for data loss, and leave very little room for little excess capacity for other workloads.
downtime. Unfortunately, data protection
With utilization on virtualized servers running so
software vendors have frequently found
high, there is less free capacity to accommodate
themselves playing catch-up to rapid changes in
backup software running on the host OS in a virtual
the virtualization operating environments.
machine. This software takes vital resources away
In the early days of virtualization, there were no from business applications running on the same
officially sanctioned methods for backup tools to machine, and can impact the performance of
interface with the hypervisors controlling virtual applications running on other virtual machines on the
machines. Without formal APIs, this led to a lot same virtual server.
of ad-hoc approaches to backup and recovery.
In addition to backup issues, early implementations
These approaches were not sanctioned by
of virtual server data protection software had
the vendors of the virtualization software and
significant recovery issues. Due to lack of
upgrades to the hypervisor inevitably broke the
granularity, restoring data to a virtual machine
backup tools, putting data protection at risk.
was frequently an all-or-nothing proposition. This
is a big problem if, for example, you only want to
restore a single corrupt file.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 7
4
4.1 Backup Issues For Small & Mid-Sized Companies
If youre a small- to mid-sized enterprises New backup solutions are often designed to
(SMEs) chances are you are at a crossroads only protect virtual servers. Many vendors
when it comes to choosing the right data of backup products were slow to respond to
protection solution for your data. Moving to virtualization requirements due to the ad-hoc
virtualized servers is not going to happen nature of the environment. This left an opening
overnight. It is highly likely that your for smaller, more nimble start-up companies.
environment has a mix of physical and The new tools these companies developed
virtual servers running a variety of business solve the specific problem of backing up virtual
applications, and will do for some time. This servers but they generally do not provide
brings up several issues. protection for existing physical servers or for
more conventional backup scenarios involving
Conventional backup tools are primarily tape.
designed to protect physical server
environments and are often not virtualization- Many SMEs will find it necessary to support
aware. Your system administrators are multiple backup tools in the short-term, but this
intimately familiar with these tools and will have will increase software and support costs and
a variety of trusted standards and procedures in reduce administrator productivity.
place that smooth management of backups and
enable swift and accurate recovery.

4.2 Agentless Backup

When thinking of backup, we need to distinguish
between the application and the machine.
Traditional backup software requires an agent be
installed on the host OS to communicate with the
backup server that catalogs and stores backup
data. Agent software helps the backup become
application-aware.

Virtualization environments are increasingly relying

on agentless backup. This approach backs up the
entire virtual machine but has less understanding
of the applications running on the host.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 8
5

New Backup
& Recovery
Capabilities
The unique nature of virtual servers has generated several opportunities for extending the
use of data protection techniques. VMware, for example, hosts each virtual machine in a core
vmdk file, with several smaller supporting files for logs, configuration, and the like. Being able to capture
an entire virtual machine by performing an image-based backup of a handful of files has its advantages.

5.1 In-Place Recovery

The vmdk file contains an encapsulated host operating further simplifying the recovery process. There are
system and the applications running within it. Restoring limitations to this technique, however.
this file to a suitable device can allow the virtual
machine to be restarted straight from the backup. This If recovering a virtual machine to a backup device,
is known as an in-place recovery. the device is unlikely to have the same capacity
and resources as the virtual server being backed
In-place recovery can get a failed server up and running up. Applications running on the device will likely
quickly. The virtual machine is restored to another have degraded performance and suffer from other
server or a backup device and restarted. This gives limitations. Also, in-place recovery invariably means
users almost immediate access to their applications and users will have to experience a second period of
data, albeit at a recovery point equivalent to the most downtime at some future time to transitioned the
recent backup. The straightforward nature of in-place virtual machine back to its original server.
recovery means that it is amenable to automation,

5.2 VM Migration
One of the benefits of server A backup of a virtual machine contains
virtualization is that virtual everything needed to set the virtual
machines are easy to migrate machine up and running for an in-
between servers. There are place recovery.
any number of reasons why you might want to do this:
moving workloads between physical and virtual servers, This also means the backup can be used as a vehicle for
or vice versa; host server maintenance and upgrades; and migrating the virtual machine, eliminating the need to
balancing workloads. schedule ad-hoc migration tasks.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 9
5
5.3 Changed Block Tracking
Hypervisor vendors have added APIs into their operating Changed Block Tracking enables backup software to
systems specifically for backup software. For VMware perform a snapshot of a virtual machine and then
the vStorage APIs for Data Protection are specifically for periodically back up only changed data. This allows
backup. VMware provide their own integrated backup for more frequent backups with much less data being
facility, VMware Data Protection (VDP), but third-party transmitted to the backup server. Recovery is shortened
vendors have also incorporate the backup APIs into their because, after a failure, only changed blocks need to be
solutions, benefiting overall stability of the backup tool restored.
and giving virtual server admins greater selection of
backup solutions to choose from.

APIs also provide greater insight into the workings of the

hypervisor, allowing for a much more granular level of
data protection. Changed Block Tracking, for example,
allows backup software to understand what has changed
since the last backup. Like traditional incremental backups,

5.4 Advanced Capabilities

Virtual machine data protection also lends itself to a number of other capabilities.

Synthetic Full Backups Multi-Hypervisor support

Synthetic full backups have been a feature of data centers Although VMware was the first modern server
for several years, but have only found a small number of virtualization solution, it is by no means the only one.
use cases. With the ability to perform periodic full backups Microsoft Hyper-V is a powerful contender in the data
of virtual machines followed by frequent changed block center and there are a number of lesser virtualization
backups, it is again possible to create synthetic fulls, which solutions from vendors such as Oracle, Red Hat, and
can be used for point-in-time recovery or archived for Citrix. It is important that any data protection solution
compliance. support multiple hypervisors. Although there are products
that lend themselves to one environment or another, the
administrative, maintenance, and productivity costs of
maintaining multiple data protection solutions should be
kept in mind.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 10
6

The Business of
Data Recovery
It is critically important to understand how
sensitive each area of your organization is to data
loss when evaluating data protection options.
This information informs technology selection, There are two industry-standard metrics
provides the foundation for your backup and used to record a business applications tolerance of
recovery and business continuity planning, and downtime and data loss: recovery point objective
lets IT know the consequences of a failure to (RPO) and recovery time objective (RTO). These
recover each business application. This is even metrics are units of time and indicate how much
more important in a virtualized setting where data application users can tolerate losing (RPO) and
an outage to a physical server can affect many how quickly an application has to be back online
different applications. before the organization begins to suffer significant
losses (RTO). RPO extends back from the time of an
outage and RTO extends forward.

6.1 Recovery Point Objective (RPO)

RPO is a measurement of data loss. The larger the RPO, the point in time that you can successfully recover data
the more data loss an application can tolerate before up to. All data between that point and the time of the
it becomes a problem for the business. Think of it as disaster will be lost.

6.2 Recovery Time Objective (RTO)

RTO is a measure of an applications importance to If you dont know the RPO and RTO of each application,
ongoing business operations. The smaller the RTO, the youre in the dark when it comes to disaster recovery.
faster you have to work to get the application back Whatever you do to ensure recovery after a disaster will
online before the organization starts to suffer losses. be guesswork. Knowing your RPO and RTO allows you
to define levels of service that you can deliver against.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 11
 7

Virtualization,
Data Protection
and Business
Continuity
The unique capabilities of backup and recovery In place recovery, for example, lets you quickly
in a virtualized environment can significantly mount a backup copy of a virtual environment on
strengthen your ability to protect application data. a different server and restart the virtual machine.
But virtualization can also create problems if you This can dramatically reduce the time it takes to
dont already have a good business continuity plan recover.
in place.
Business continuity is rarely a straightforward
When employees work around the clock and matter of restoring a single virtual machine. Inter-
business is effectively always on, business dependencies between applications on different
continuity planning is critical. Any disruption to virtual machines and between virtual and physical
normal operations can quickly lead to lost revenue, servers can introduce complications. Avoiding
lost productivity, lost brand value, and potential the following five common missteps will help
compliance issues. you maintain perspective in planning for business
continuity.
Virtualization can enhance your ability to get mission-
critical systems back online after an outage.

7.1 Its About The Business

Disaster recovery, high availability, backup and Before trying to work out how to implement disaster
recovery, business continuity, call it what you will, recovery, you need to spend time thinking about:
the aim is the same: keep the business up and Why? Talk to business leaders to understand their
running no matter what the circumstances. Too priorities. For some it will be email, for others the online
often, organizations let technology take the lead and order entry system, for others Microsoft SharePoint. The
dominate the conversation. What is often forgotten, point is, you wont know what systems are the most
and is essential to remember, is that disaster recovery important unless you ask business users. Understanding
is about satisfying a business need. It must be driven by the needs of the organization will let you set priorities
business requirements. that dictate your disaster recovery technology choices.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 12
7
7.2 Downtime Costs 7.4 Workable Plans
Too often, organizations assign a dollar value for If your disaster recovery plan is a Post-It note on the
disaster recovery planning before analyzing the backup tapes under your system admins bed, youre in
financial risk of downtime and data loss to the trouble. As crazy as it sounds, a surprising number of
business. If you cant quantify how much you can lose organizations dont have a disaster recovery plan. It is
from an outage to critical systems, it will be difficult to essential that you develop a formal document detailing
know how much you can spend to avoid these losses. all applications, hardware, facilities, service providers,
personnel, and priorities.
Recovery point objectives (RPOs) and recovery time
objectives (RTOs) will help you understand how Maintaining a disaster recovery plan is only helpful if it
sensitive each area of your business is to downtime works, and the only way to ensure that your plan works
and data loss. Knowing the RPO and RTO for each is to test it. Testing the plan under simulated disaster
business application will help you calculate the cost of conditions can be challenging, and time-consuming.
downtime and enable you to define levels of service Fortunately, todays leading data protection solutions
that you can deliver against. provide the ability to automate testing of your disaster
recovery preparedness. This can now be done without
downtime and without taking production applications
7.3 Measuring Risk offline. Look for data protection solutions that help you
create environments for non-disruptive testing of your
Exactly what events classify as a disaster will disaster recovery plan.
vary from organization to organization, and even
from department to department. Some events
earthquakes, for exampleare potentially so
catastrophic that it is obvious the organization
7.5 Fallback
must protect itself against their occurrence. Other Moving systems back to the production environment
events may be more commonsuch as faulty after failing over to a disaster site is an often overlooked
network hardwareyet have an outsized financial component of disaster recovery planning. Its easy to see
impact. When thinking about disaster recovery, it why. When we think of disaster, our minds focus solely
is essential to ask: What are we trying to protect on protecting valuable assets. Little thought is given to
ourselves from? Dont overlook the what happens to those assets after the disaster event has
commonplace. Small losses passed.
from common problems can
mount up quickly. The ability to failback to production systems is every bit
as important as the ability to failover. Unless carefully
planned, a backup data center is unlikely to have the
same capacity or performance as the production
site. Without a fallback plan, you may perform
a successful initial failover and then see
losses mount as your business limps along
for weeks operating from an inadequately
provisioned backup site.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 13
8

Virtualization
and Tape
Tape has been a reliable medium for storing production
backups and archives for decades and it continues to play
an integral role in the backup strategies of many companies.
Disk-based data replication and the cloud have reduced demand
for tape as a tier-one backup target, but the benefits of tape are
such that it continues to have a role to play. Its critical to include tape
backup capability in the evaluation of virtual server backup and recovery
solutions, especially if existing physical servers rely on tape backup.

8.1 Industry Standards 8.3 Portability

One of the benefits of a technology with a long track Cloud storage can offer definite friction-free benefits when
record is industry standardization. Linear Tape-Open storing production data offsite, but its not appropriate for all
(LTO) has proven to be an enduring format, with applications, or all organizations. For example, its not unusual
the most recent iteration of the standard supporting for regulators to demand corporate archives be stored in the
up to 2.5 TB of raw data per cartridge and transfer country of origin. This can be difficult to guarantee when backup
speeds of 160 MB per second. These standards are data is hosted in a third-party service providers data center
still evolving. Future LTO cartridges are anticipated to somewhere in the cloud. Backup data volumes can also prohibit
store up to 48TB (120TB compressed) and transfer network transfers to cloud storage providers. And, as Andrew S.
data at terabyte-per-second speeds. For large Tanenbaum observed, Never underestimate the bandwidth of a
datasets, the capacity and speed benefits of tape station wagon full of tapes hurtling down the highway.
dwarf alternatives.

8.4 Good Enough for Google

8.2 Cost and Durability In 2011, Googles Gmail service experienced a very public
Magnetic tape is an exceptionally cost-effective outage. A bug introduced in a new software release managed
and durable medium. Bit Error Rate (BER), mean to propagate through multiple data centers before it was
time between failure (MTBF) rate, and bit rot spotted. Gmail users woke up to empty mailboxes and Google
the gradual decay of data stored on magnetic did not have the option to failover, because backup data
mediaare all lower for tape, compared to disk, and centers had the same problem. Google eventually resolved the
manufacturers routinely quote a 30-year lifespan for problem by restoring user mailboxes from tape backup. Many
tape cartridges. It is undoubtedly the safest medium industry observers were surprised by Googles use of tape, but
for long-term data storage. it highlights the versatility of the medium. Organizations large
and small can learn from Googles experience of recovering
from the Gmail outage and apply those lessons to their own
data protection and business continuity plans.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 14
9

Data Protection
and the Hybrid
Data Center
Todays data centers are more often than not Ensuring the availability and integrity of data in
a hybrid, with IT architects selecting the most this new hybrid data center is a challenge. Legacy
appropriate infrastructure for business applications backup and recovery tools may be great for physical
based on a list of proven best-in-class alternatives. servers and tape, but not so good with virtual
The cloud now offers an abundance of off-the- servers. New software for backing up virtual servers
shelf business processes available as Software- may not work well on physical servers, and almost
as-a-Service (SaaS) offerings, for fast deployment certainly wont be robust enough for tape. With
and pay-as-you-go economy. Virtualization-first IT budgets and staffing at a standstill, its more
policies frequently dictate the host for new home- important than ever to avoid the trap of having
spun applications and packaged software solutions silos of data protection software for each server
that must be run from the safety of the corporate architecture. This can quickly raise costs and sap
data center. And, physical servers continue to play administrator productivity.
a role for legacy applications and those processes
that demand the highest performance or specific
hardware tie-ins.

9.1 When is HA Not HA?

High availability (HA) technology is now a very real option As Google discovered during their Gmail outage, there is
for the mission-critical apps of mid-market companies. a role for conventional backup even when you have an
The technology is no longer the complex, esoteric HA configuration up and running. However, there should
approach to business continuity that it once was. be no confusing conventional backup techniques for HA.

The goal of HA is simple: zero downtime and zero data

loss. Tools do exist that lay claim to being HA, but if they
are not able to eliminate your exposure to downtime and
data loss, theyre not HA.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 15
10

Arcserve Unified
Data Protection
Hybrid data center architectures are complex enough without adding the
additional risk of incompatible backup and recovery solutions. However, this
is often the path organizations take when deploying a combination of legacy
tools and newer single-purpose data protection technologies. The lack of
integration inevitably creates silos that add unwanted management, increase
overhead costs in the infrastructure, and complicate recovery - effectively
working against the objective of reducing risks to operational data.

Arcserve Unified Data Protection

(UDP) is a single, unified data
protection solution that provides
the flexibility to support a wide
variety physical, virtual, and cloud
IT platforms, a diverse range
of application RPO and RTO
requirements, and an array of
backup media, including tape.

Arcserve UDP integrates disk-to-disk

backup, tape backup, replication,
high availability, and global
deduplication in a highly scalable,
single architecture. With agentless
deep integration to support a
variety of hypervisors and a modern
task-based approach to administration, Arcserve UDP is able to automate
complex repetitive tasks and provide all data protection and recovery from
a single pane of glass management console. Assured Recovery, a unique
capability of Arcserve UDP, offers automated, risk-free testing of disaster
recovery scenarios, without requiring end-user downtime.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 16
10

10.1 Arcserve UDP and Magnetic Tape

Arcserve UDP offers enterprise-level functionality
to give small- and mid-sized organizations a single
tool for their hybrid IT infrastructures. The solution
offers bare metal recovery (BMR), local and remote
standby, instant VM recovery, push-button failover
and failback, and granular recovery from any backup
media, including tape.

Unlike point solutions, Arcserve UDPs support

for magnetic tape is not a last minute add-on.
With a 25-year history providing data protection
for organizations and enterprises of all sizes,
Arcserve has deep experience supporting tape.
Tape-awareness is built into the core of the UDP
10.2 Arcserve UDP 7000
technology. This is critical when integrating tape
with modern data protection technologies like
Appliance Series
global deduplication, incremental backup, and VM The Arcserve UDP 7000 Appliance series provides
snapshots. a set and forget backup and recovery solution in
a cost-effective, purpose-built data protection
solution. The hardware easily integrates into existing
data protection schemes and offers cloud-native
capabilities, unmatched ease of deployment and use,
and output to magnetic tape.

Arcserve UDP 7000 Appliances support global

deduplication, multi-site replication, tape backup, and
automated data recovery. Further, they seamlessly
integrate with Arcserve UDP software to provide
a distributed data protection architecture. Global
deduplication technology actively reduces backup and
bandwidth costs by combining proprietary, industry-
leading deduplication, coupled with incremental
backup technology to reduce disk, tape, and network
capacity needs. Together, these capabilities increase
operational agility and simplify disaster recovery.

Copyright 2016 Arcserve (USA), LLC and its affiliates and subsidiaries. All rights reserved. Arcserve | 17