Cloud Computing Hosting and Servers are gaining traction with small and medium business at a breakneck pace.
Every day, we are sold on the idea of lower cost, better up-time, ease of implementation and more.
However, most people are unaware what a true cloud really is and consequently they may end up choosing a
CINO hosting provider or in other words a Cloud In Name Only provider.
Cloud – the big picture view
The essential way to think of a cloud is a like a ‘utility’ such as water, electric, telephone, natural gas, and so on. As with any service, you use what you need and pay accordingly.
For instance our ‘need’ (resource consumption) for water goes up greatly in the summer for trees and lawns and to stay cool. Our expectation is that we turn on the tap and consume the amount of water we deem fit. When done, we turn it off. Our water company will read the ‘meter’ and bill us for the water consumed. Water companies don’t have a practice of saying you must buy thousands of more gallons of water than you’ll need. They simply ensure the infrastructure is working and able to bring the water to the tap, and that it’s safe for human consumption (secure) and available 24x7x365 (high availability).
Cloud hosting works in much the same way. Your website or application will consume a certain amount of resources such memory, disk storage, CPU or network. You would be billed according to the amount of resource consumed.
Features of a Cloud Server
The national institute of standards (NIST) has defined what makes up a cloud-computing platform. While the actual method of implementation is left up to the vendor, a true cloud server implementation will offer the following features.
On-Demand Self Service
Cloud computing by its nature is natively self-serve, allowing the consumer to provision their own needs, such as the amount of CPU, RAM, and network storage (SAN or NAS), as needed without requiring human intervention from the host. As an example a ‘cloud’ that requires migration of your data to a new machine to upgrade is not on-demand.
Broad Network Access
This is the idea of ubiquitous connectivity to most any device type (desktop, phones, tablets, mobile computing, etc.) and typically will be more applicable to a Software-As-A-Service (SaaS) Services delivery Model. This feature specifically is related to such things as hosted online learning education Software or other software and applications, where your ‘application’ resides in the cloud.
One of the most attractive features of the cloud hosting model is resource pooling. Pooling is designed to work by taking multiple servers (including their sub-components such as memory, cpu, bandwidth, network storage) and pooling them into a common set of resources. Resource pooling allows for multi-tenant’ model that allows access to different physical and virtual resources as required by the consumer need.
Cloud hosting infrastructures are elastic. They can have resources added or removed very rapidly depending on the demand (load on the website or application). This gives the consumer websites for instance, a “view” of nearly unlimited resources.
Cloud hosting gives has the native ability to provide detailed resource measurement. This is accomplished through the measurement at various points in the infrastructure for usage of such as memory used, disk storage consumed, bandwidth usage and so on. This gives the consumer a transparent billing model to the consumer.
Cloud hosting services are delivered vis-à-vis a utility model being offered in three different variations.
Software as a Service (SaaS)
SaaS offers the consumer access to ‘applications’ running on a cloud infrastructure. The consumer can access it through various devices, and typically buys on a subscription basis. The consumer has no control over the cloud infrastructure.
Platform as a Service (PaaS)
In this service model the consumer can deploy their own apps or websites, such as those built on a content management framework (Joomla!®, WordPress®, Drupal, Magento, etc). Billing is typically pay-per-use or charge per use basis – such as consumption of service (remember clouds are a UTILITY model).
The cloud itself consists of hardware and software that conforms to the essential characteristics of cloud computing. Keeping with the spirit of cloud computing, the consumer does not manage the cloud, the network services, operating systems and so forth.
Infrastructure as a Service (IaaS)
In the final model, the consumer has a greater amount of control, such as the selection of the operating system and applications. Yet again they do not manage the cloud itself. In this model the consumer has, [provider dependent], limited ability to handle some network components such as firewalls
How Rochen does it
With Rochen Managed Cloud Servers (MCS) you are in control of scaling of your resources such as the memory, CPU and enterprise grade SAN. Additionally with our Cisco network infrastructure, you can rest assured that the power and bandwidth you need is there.
With Rochen’s MCS you have three methods of increasing your resources, all in a powerful, self serve manner:
Using ‘Real-time Manual Scaling’ your cloud server instance can be increased on demand. Automatic scaling gives you the comfort that your site won’t run out of resources. This feature increases resources as they are requested. The last method is Scheduled Scaling. This would be used in the case of an upcoming season such as the Christmas shopping season, or a promotion you might be running. All of these handle the billing automatically for you – no waiting on someone to provision.Figure 2: IMMEDIATE scaling in Rochen MCS
MCS is built upon the leader in virtualization technology, VMWare®, ensuring you that in the event of a failure of a hypervisor another is already there taking up its load. Our redundant Cisco Network infrastructure offers redundant router cores to protect your connection to the Internet. In fact in the Managed Cloud Server service, there is NO single point of failure giving you redundancy and security.Figure 3 – Managed Cloud Server’s – no single point of failure
In addition to Rochen’s 24x7x365 Global support engineers monitoring efforts, MCS gives you an added layer of assurance, in our ‘single-pane-of glass’ dashboard. This gives you a single place to view your scaling needs, monitor resources, and see open tickets requiring your attention, billing information, technical knowledge resources and more.
Figure 4 Current Services status
Figure 5 CPU Usage for an example MCS instance
In the previous figure you see the three areas of resources you can monitor, CPU, RAM and disk. This gives you the knowledge you need to make a determination if you need more or less resource. This real time analytic is important for reviewing amount of resources in use. If you see you’re peaking – you can add more manually or via Auto Scaling.
Auto ScalingFigure 6 – auto scale settings
In this image, the scaling up for your storage, RAM or CPU is easily done independently of each other. The ability to set them by increments is built right in as shown. You can specify which of these you would like to increment up and by how much. In the event of a particular event coming up such as a promotion, holiday sales or other traffic generating activities, save yourself the hassle of downtime or slow servers and schedule ahead.
Calendar Based Scaling
Since this is written in mid November, it makes me think that Christmas time is right around the corner. With the expected increase in site traffic, you might expect, with calendar scaling you can be ready. Simply set the date and time you want to scale, the resources you want to scale and click save. Rochen Managed Cloud Servers are Very ‘fire and forget’ in its design.
Rochen Managed Cloud Servers offer you a ‘true cloud’ hosting experience. With no single points of failure, Manage Cloud Servers are the GO-TO choice for highly available, highly scalable cloud servers.
Tom Canavan is a member of Rochen’s ‘Enterprise Solutions Group’ focused on creating customized and complex hosting solutions.