Building a Private Cloud, Step by Step
The private cloud is a well-used architecture and technology within enterprises. Even so, most enterprises are still not well-versed in how to deﬁne, build, and deploy private clouds. There is a “gap of confusion” that exists between the proper use of private clouds and rank-and-ﬁle enterprise IT.
In this page we’ll attempt to unveil the mystery of private cloud. We will provide you with a step-by-step guide to understand your own requirements, pick the right hardware and software, deal with secu- rity and governance, and ﬁnally implement the private cloud solution that will get it right the ﬁrst time.
Why Private Cloud
Private and hybrid clouds are emerging around common use case patterns that are most likely to provide value for enterprises. They include:
Deploying to private and hybrid clouds to deal with industry regulations and, ultimately, risk. For instance, PII (Personally Identiﬁable Information), if breached, can cause all a legal and public relations nightmare. If the cost of this risk outweighs the additional cost of using a private cloud, then the private cloud should be selected. Other examples include control and security around ﬁnancial information, as well as some customer data.
Deploying to private or hybrid clouds in order to have dedicated hardware resources to ensure high and consistent performance of critical systems. These include large database systems that need direct access to underlying hardware systems and fast I/O, which won’t function well with “bursty” performance metrics on a public cloud. Typical use cases include predictive analytics, consumer-facing transactional systems, HPC, and even gaming – many gaming platforms are moving in-house to control cost.
Using the private cloud to save money. There are two common scenarios: First, a sunk cost in in-house hardware and software systems, which can be leveraged for a private cloud, instead of spending more on public clouds. Second, when the cost of public cloud resources exceeds that of maintaining a private or hybrid cloud infrastructure. Despite the hype, at a certain scale public clouds will cost more than private clouds, when considering the costs holistically.
Using private or hybrid cloud to support an emerging DevOps organization. Public clouds are often the desired end-point, when it comes to DevOps and application development that can be automated with DevOps approaches and tools.
Private Clouds – Big Benefits, Big Complexity
The private cloud oﬀers most of the advantages of the public cloud: Self-service and scalability, multi-tenancy, the ability to provision machines and computing resources on-demand, compliance, and improved security.
However, complexity issues include:
• Time-consuming infrastructure conﬁguration and maintenance.
• Inconsistent visibility into infrastructure utilization and cost.
• Ever increasing complexity with custom requirements in every project.
Building a Private Cloud
Now let’s look at what it takes, step by step, to build a private cloud. While there are many patterns that can be found in traditional IT, there are some new approaches and technology that must be understood.
Step 1: Deﬁne the Purpose: Understand the requirements of the business and those force on you by security regulations and operational considerations.
Step 2: Deﬁne the Workloads: Determine what types of applications and application data will run on the private cloud, by dividing workloads into applications, data and infrastructure.
Step 3: Deﬁne the Hardware: Take the data gathered in the previous step and size up a hard- ware system that will provide the right support now, and into the future
Step 4: Deﬁne the Software: Decide if you want to go proprietary or open. If you’re moving to open solutions, OpenStack is the primary choice.
Step 5: Deﬁne the Network: Define how your network will work on your private cloud – physical network provisioning, software defined network elements if any, security, network management.
Step 6: Deﬁne Security: Plan your Identity and Access Management (IAM) – a security approach and technology that enables the right individuals to access the right resources, at the right times.
Step 7: Deﬁne Governance: Once you get to a certain number of cloud services, you won’t be able to keep track of them all and provide the control they will require. Plan your service governance model in advance.
Step 8: Deﬁne Management Processes and Tools: Define monitoring practices, physical infrastructure including network, power supplies, and more.
Step 9: Implementation: Stage the private cloud, including hardware and software in your data center.
Step 10: Testing: Define test processes to verify you are fully meeting requirements and prepared for security issues, downtime or failure.
Step 11: Operations: Deﬁne how you will operate the cloud – monitoring, automation, security, governance, etc. This is known as CloudOps.