Global Server Load Balancing (GSLB) enables multi-data center and multi-cloud resilience by leveraging service resource awareness and DNS to steer traffic across geographically distributed pools based on defined business logic. When deployed across multiple data centers, it provides application load balancing at geographic scale – across town or across the globe – using any mix of hardware, virtual and cloud platforms. GSLB provides multi-site resilience with seamless failover and failback in the event of a critical resource failure as well as offering optimised redirection of traffic to the closest physical service location. In the event of a service disruption, traffic is automatically redirected based on predefined policies to minimize impact and the need for manual intervention.
Client traffic is sent to the location that will provide the very best application performance and client experience, tailored to the location of the client and the observed availability of each location. The client traffic redirection can be further refined using EDNS (Extension Mechanisms for DNS) to provide fine-grained control on which available locations a client is directed based on their source subnet.
Providing these services locally is a cost-effective alternative to providing these services using in-house resources or contracting with expensive hosted Domain Name Service (DNS) services that don’t provide the flexibility and control that you require. Services like company email, remote desktop environments, and other managed resources where user experience, high availability, and disaster recovery are mission critical.
In the classic GSLB use case, GEO is deployed in an Active-Active configuration to provide high availability and optimal user experience across multiple geographically distant endpoints.
Sign up here to watch an exclusive Global Server Load Balancing (GSLB) Webinar delivered by our Principal Technical Advisor, Ben Hodge.
Clients from around the world attempt to connect to services that are hosted in three geographically distant locations: the USA, the EU, and Australia. Kemp GEO appliances running at each location cooperate to decide which data center will provide the “best” user experience for a client connecting from a particular location – where “best” is defined by the parameters that you configure into your GEO configuration.
This is often called an Active-Active configuration -- all the endpoints operate across multiple sites simultaneously and have access to required back end data. Here GEO actively measures the real time load at each participating site and employs predictive analytics to ensure that requests are distributed for optimal performance and resources across the multi-site infrastructure. Additional logic may also be implemented to steer specific clients to specific sites depending on knowledge of network performance as well (such as round-trip communication times to specific locations).
GEO is also an effective Disaster Recovery (DR) solution, where multiple data centers are deployed in an Active-Passive configuration.
An alternate data center or cloud infrastructure is used as a backup to the primary production environment. The primary location is actively serving traffic, while the backup is idle and in a standby state. GEO determines when the primary site has failed as well as when services in the secondary location have become operational, only directing traffic to that site once required services at that site are ready and can provide a smooth transition of client services.
Finally, GEO can also help you cost-effectively manage expansion into public clouds to provide on-demand service expansion during high volume periods, which uses elements of both Active-Active and Active-Passive configurations.
In the simplest case, there might be one off-premises datacenter location that can handle all application traffic 80% of the time. Let’s say there this location runs all its applications and infrastructure on a combination of hardware and HyperV. There is also a second datacenter located in a public cloud such as Amazon Web services (AWS) or Azure, which is on standby as in an Active-Passive configuration.
When the traffic rises to a certain load level on the primary datacenter, GEO begins to direct traffic to the second datacenter as in an Active-Active configuration, essentially bursting out additional capacity to handle the additional traffic. Once the traffic flow returns to a level that the first datacenter can manage on its own, GEO stops directing traffic to the second datacenter.
Since the second datacenter is located in a public cloud, GEO is helping to provide the best possible user experience while keeping service costs down by directing traffic to the public cloud only when the level of client traffic requires it.
The Harris County District Attorney’s Office required an intelligent load balancing solution with Geographic Server Load Balancing (GSLB) functionality, to provide high availability for its mission critical applications and multi-resiliency across its production and disaster recovery site in the event of a critical resource failure…. Read more
Regardless of how you deploy, you can be assured that Kemp’s integrated GEO global server load balancing solution and LoadMaster local application delivery traffic management and load balancing will help you optimize and protect your application infrastructure with an easy-to-use and cost-effective solution.
GEO is available as standard with your support subscription on all LoadMaster load balancers across all hardware and virtual models and platforms. GEO is also available as a standalone virtual appliance.