UNIT XI: Private Clouds:
Understanding Key Private Cloud Requirements –
On-demand self-service
A consumer can unilaterally provision computing capabilities, such as server time and
network storage, as needed, automatically, without requiring human interaction with
each service provider.
Broad network access
Capabilities are available over the network and accessed through standard mechanisms
that promote use by heterogeneous thin or thick client platforms (e.g., mobile phones,
laptops, and PDAs).
Resource pooling
The provider’s computing resources are pooled to serve multiple consumers using
a multitenant model, with different physical and virtual resources dynamically
assigned and reassigned according to consumer demand. There is a sense of location
independence in that the customer generally has no control or knowledge over
the exact location of the provided resources, but may be able to specify location at
a higher level of abstraction.
Rapid elasticity
Capabilities can be rapidly and elastically provisioned, in some cases automatically,
to quickly scale out and rapidly released to quickly scale in. To the consumer,
the capabilities available for provisioning often appear to be unlimited and can be
purchased in any quantity at any time.
Measured service
Cloud systems automatically control and optimize resource use by leveraging
a metering capability at some level of abstraction appropriate to the type of service
(e.g., storage, processing, bandwidth, and active user accounts). Resource usage can
be monitored, controlled, and reported providing transparency for both the provider
and consumer of the utilized service.
Corporations worldwide are adopting cloud computing as the main computing
platform for conducting their day-to-day business. They have abandoned or severely
cut back on capital expenditures and maintenance for on-site servers, storage
devices, business applications and backup devices such as tape and optical drives.
By eliminating hardware or software acquisition and maintenance costs, companies
realize a substantial savings when moving to a cloud-based architecture.
While the savings can be promising, many corporate IT directors have valid concerns
about the cloud computing trend, including that of entrusting key corporate assets to
a new way of deployment that they are not completely familiar with, and comfortable
implementing on a company-wide basis.
A complete examination of cloud computing strengths and weaknesses allows
corporate IT managers to better determine whether cloud computing is a viable
option for all, or select parts, of their business needs.
Gap Analysis of Existing Infrastructure and Processes –
Please refer to link.
Migration Process to Private Cloud Computing –
Once a business has decided it can commit to a cloud computing infrastructure, it can begin to
decide how it will migrate to the cloud. A number of considerations come into play, such as the size and complexity of current applications, the amount and number of applications, and whether the existing legacy platform is overdue for upgrades or replacements.
Incremental Cloud Deployment
Most companies start out incrementally, choosing a SaaS or IaaS application that
would replace an in-house application, such as CRM, Salesforce automation, or email.
After the new application settles into the business process, another cloud-based
application can be added. In this incremental approach it is also advisable to choose
the least disruptive application changeover first, or to migrate less business-sensitive
applications earlier. It is in the best interest of the company to carefully document the
migration of each system or application and also, if possible, to capture the difference
between cost of premises ownership versus cost of cloud implementation.
Legacy Migration
With very entrenched legacy applications, equipment, and computing processes,
it may take time to migrate every application from one platform to the next, especially
when application performance and reliability are at stake. It is imperative that
businesses with a large base of legacy computing applications have a competent and
experienced cloud computing provider to help them reengineer their systems for
optimum performance and reliability.
Productivity, customer satisfaction, and business continuity cannot suffer during
a migration. Companies should create mirror systems of key legacy applications —
one on the new cloud platform, another on the existing platform, and compare
performance, reliability, functionality before cutting over to the cloud-based version.
Also, companies should consider moving the most critical legacy applications last.
Avoiding Costly Upgrades
When businesses are faced with costly and complicated upgrades to applications, it
may be the right time to consider cloud computing alternatives. Rather than pay for
costly computer, software, and storage upgrades and potentially greater outlays for
hardware maintenance, companies facing upgrades can upload existing data from
legacy applications to cloud computing alternatives, test the applications on live data,
and then retire the legacy application once the platform has been fully tested.
Straight to the Cloud
Customers with relatively simple legacy applications, such as business productivity
software and a few custom applications, are better off making a complete changeover.
The cloud infrastructure and applications can be configured, tested, and rolled out in
tandem with old systems, and cut over can occur fairly quickly. Again, an experienced
cloud provider is essential in making a smooth transition.
Technology and Vendor Options –
| Feature | Eucalyptus | OpenStack | CloudStack |
| EC2 Query API |  |
|
|
| EC2 Soap API | |
 |
|
| OpenStack API / Rackspace API | |
|
|
| SSL Between Components | |
|
|
| Horizontal Component Scalability | |
|
|
| Web-based UI | |
|
|
| Command line interface | |
|
|
|
Type
|
Weakness
|
Strength
|
|
Openstack
|
– Young codebase- Uncertain future
– Initial configuration |
– Single codebase- Growing community
– Corporate support |
|
Eucalyptus
|
– Install requirements- Configurable but not very customizable – Community Inclusion |
– Excellent commercial support- Fault-tolerance
– Offers a hybrid-cloud solution with AWS |
|
CloudStack
|
– Very GUI centric- Single java core
– AWS integration weak |
– Well-rounded GUI- Stack is fairly simple
– Customization of the storage backend |
|
Ganeti
|
– Admin centric- VM Deployment
– No AWS integration |
– Fault-tolerance built-in- Customizable
– Very simple to manage and maintain |
Key Challenges in Creating a Private Cloud.
Most of these potential fears involve security, availability, flexibility and control, quality
of service, or the long-term protection of valuable data. Others are concerned that
the connection to the off-site cloud will not perform consistently or that their entire
business would rely on network connections that may go down at any time. Every
company should fully address these issues, and understand the pros and cons, before
migrating specific IT functions and applications to a cloud platform.
Cloud Security
Companies considering cloud computing should look carefully at the security and
reliability of potential service providers. Most cloud computing vendors make security
a top priority. Since many service providers began their businesses as online commerce
providers, they have ample experience, tools, and equipment in securing online data from
unauthorized use and in making certain that access to that data is highly secured as
well. Cloud computing providers offering SaaS and IaaS services routinely ensure that
company data is segregated through firewall and password access.
Other layers of protection for businesses using cloud computing are private network or
VPN and/or IPSec connections to their cloud. This technology ensures that the entire
connection from the off-site worker’s desktop to the cloud network is safe from data
capture or interception of any kind. Some providers use Multiprotocol Label Switching (MPLS) networks that can route traffic dynamically. These are highly reliable and highly secure
High Availability
The most complete cloud computing facilities are routinely protected from equipment
failures and outages using redundant network switches, servers and storage facilities.
In particular, the use of off-site backup and redundant servers and storage facilities
make these well-equipped cloud computing facilities less vulnerable to disaster, local
catastrophe or malicious attack.
The best cloud computing providers also see to it that customers can have fully
redundant IP connections to the cloud, protecting access in the event that a local
disaster or equipment failure drops carrier access to the cloud platform. Using MPLS
also adds an additional layer of reliability and performance to cloud access.
Guaranteed Service Levels
Rightfully, corporations should be concerned with the level of service that cloud computing providers can actually guarantee. In most cases, these vendors can only control SLAs on their cloud platform, leaving the service level of access to the cloud to a third-party service
provider. Therefore, it is important that businesses get SLA guarantees from this provider from the company environment/desktop to their cloud provider to ensure that the business has the best possible access and uptime for services in the cloud. If the cloud is up and running, but
access to it is down due to an unreliable connectivity provider, the entire business may as well be closed.
While having SLAs is important, it is also important to define how and when an SLA truly goes “into effect.” An SLA does not necessarily affect the actual operations; its terms and conditions are only recited when things go awry, and it typically does not protect a business from loss of system uptime. Similarly, an SLA will not prove an effective tool for managing a cloud provider. It will still be a prerequisite, and the SLA offered by the cloud provider under consideration should be consistent with what others provide.
Is uptime exclusive of scheduled maintenance? Is there a cap to service credit? While payment should cover the cost of the lost service, do not expect business losses owing to the downtime
to be covered. It is also important to understand not just the contractual entitlements, but the course of action to seek a remedy. Does the provider offer a 24-hour help line?
In the end, the quality of the uptime is directly related to the sophistication of the IT department, not the strength of the SLA. Choosing an IaaS or SaaS provider that employs best practices
in design and operations and promotes transparency offers the greatest assurance of performance.
Flexibility
Most companies are concerned about the flexibility and response times associated
with change requests that may come in. Since most migrations to the cloud today are
in a shared environment, there will be some flexibility to get the full business/financial
advantages, and as cloud computing continues to mature, cloud computing companies
can offer even greater flexibility to their customers.
Managing a Cloud Infrastructure
Many corporate IT managers are reluctant to commit to cloud computing because
it also commits them to managing the new cloud infrastructure. While this may suit
some IT departments, most companies are better served moving to a fully managed
cloud environment.
Fully managed cloud environments should monitor performance, application
availability, routine backup and archiving, storage, memory and server provisioning,
and other security, performance and protections tasks. For most companies, 24X7
support and technical assistance is mandatory. Companies should read the fine print to
make certain the level of management, service, and support they require are provided
by the cloud vendor, and whether these are part of the paid service or layered on as
extra charges to the company.
An Escape Hatch
A full commitment to cloud computing often seems risky to any company. Many CIOs
fear that if the platform does not work out, it could literally ruin their credibility.
For this reason, companies should make certain that any potential cloud computing
partner has a fully documented path for not only migrating applications and systems
on to their platform, but also has defined policies and procedures, in the event that a
customer wants to disengage specific deployments from the cloud provider.
Cloud Computing 