Cloud Load Balancing
The Hybrid Cloud
In today’s complex and rapidly changing business environment, agility, and a relentless drive down on costs, are seen as paramount in data networking and application provision. For this reason, many enterprise and government organizations are turning to cloud-based data environments, with Cisco® predicting that by 2018, 78% of data centre workloads will be being processed in the cloud.
The objective is to maximise the performance of websites and applications, providing a reliable and seamless experience for users and customers, whilst retaining flexibility and minimizing overall costs of delivery. Energy consumption is also increasingly a factor in data centre provisioning decisions.
A major factor in the migration towards cloud technologies is the depth of virtualization offered, which allows dynamic deployment of workloads to meet the fluctuating demands of users and customers. In order to meet these varying needs, organizations are increasingly adopting what is called the hybrid cloud model, a powerful mixture of their own data centres and private cloud facilities, combined with the services of a public cloud operator such as Amazon® Web Services (AWS), Microsoft® Azure™, Google® Cloud Platform, Rackspace® or VMware vCloud® Air. For example, a company may retain its employee database within an in-house data centre, but migrate e-mail software, such as Microsoft Exchange Server, to a public cloud.
The hybrid cloud offers maximum flexibility, opening up options in virtualization, scalability and sophisticated application delivery. For example, many organizations plan to cope with temporary, unexpected spikes in application demand by automatically pulling in public cloud resources as required, a process known as ‘cloudbursting’. This technique can also be employed in disaster recovery strategies. Other organizations operate a bi-modal model of IT delivery in which the public cloud is used for exploratory business developments where agility and rapid implementation are key, leaving private and internal data centres to support existing, core, business systems.
1 Cisco, Cisco Global Cloud Index: Forecast and Methodology, 2013–2018, (San Jose, CA: Cisco, 2014), http://www.cisco.com/c/en/us/solutions/collateral/service-provider/global-cloud-index-gci/Cloud_Index_White_Paper.html
2 Private cloud facilities are modelled on the ideas and open technologies of public clouds but are proprietary solutions, managed by the organization themselves, or third parties, and operated behind the perimeter firewall.
3 Gartner, Gartner Says Worldwide Cloud Infrastructure-as-a-Service Spending to Grow 32.8 Percent in 2015, (Sydney: Gartner, 18 May 2015), http://www.gartner.com/newsroom/id/3055225
Cloud Load Balancing
In such a complex operational environment the use of advanced load balancer technology – Cloud Load Balancing – to orchestrate resources and workloads is critical.
Cloud Load Balancers manage online traffic by carefully distributing workloads, data traffic, and application requests across the different types of server deployment. Their most important function is to handle a variant of traditional global server load balancing (GSLB), which determines the best server location to deliver an application for an individual user or customer, and to extend this functionality to the cloud.
GSLB relies on automated decision-making algorithms that take account of the user’s location, the access device technology they are using, the time of day, application response times, and data centre and cloud availability and capacity. Based on these decisions DNS requests are sent to the most appropriate server, whether located in an in-house data centre, private cloud or public cloud. These load balancers also undertake the standard balancing techniques such as: layer 4/7 algorithms, session persistence, HTTP caching, TCP pooling, SSL offloading and DDoS attack security.
Cloud load balancing technologies
Cloud load balancing technologies come in two distinct varieties. Firstly, there are standard, hardware-based load balancers or Application Delivery Controllers (ADCs) that have additional functionality for use in a private cloud environment. Virtualized versions of these types of load balancers and ADCs are also available for use in bare metal, hypervisor environments such as Microsoft Hyper-V® and VMware ESXi™. These software-based load balancers help support the development of private clouds, operating in concert with traditional load balancers and ADCs based within the in-house data centre.
Secondly, there are virtualized versions of these load balancers which can operate as virtual machine images within public cloud systems (for example, as an Amazon Machine Image (AMI) within AWS). By deploying an application together with an associated virtualized ADC (vADC) or load balancer in a public cloud, organizations can remotely manage the cloud-based balancing for the application. Many of these software-based devices are designed to work with cloud orchestration systems such as OpenStack® and VMware vCloud Director™.
How it Works
activereach cloud load balancers for UK business
Cloud load balancing is a complex technology and deployment in a mixed environment of private and public cloud can be challenging. It is, however, becoming a key component in the on-going battle for business continuity and network performance. activereach offers a range of consultancy, architectural design, equipment deployment and support solutions for organisations looking to move into, or expand, their cloud deployments.
Whether it be public, private or hybrid cloud, virtualised or hardware-based, activereach has the expertise to help UK organisations take their data centre deployments securely to the next stage.