Availability by Load Distribution
The availability and scalability of the service are substantially enhanced by using several parallel instances of an application on separate servers and providing an upstream hardware load balancer.
Distributing the Workload
We employ load balancing solutions to distribute the load over several front end servers even in high load situations. The load balancers represent the first level of request processing in a server farm. The load balancer queries the status of all redundant application servers in real time. By means of an active feedback system, the load balancer can evenly distribute load peaks over all available systems. At the same time, load balancers provide the flexibility of expanding the server farm with any number of servers or instances in a computing cloud.
High-Availability Around the Clock
If a server fails or is being serviced, the load balancer immediately removes it from the active operating process. The result for your online service: High-availability around-the-clock 24/7/365.
The layer 7 functionality of the load balancer enables availability to be checked in real time extending up to the application level (layer 7 health check) and assigns a session (session stickiness) such as a customer's online shopping cart to the physical instance.
The active load balancer is the central hub of a server farm. As with all other systems in our high-availability environments, the load balancer is always redundant and is never set up individually. The passive load balancer immediately takes over the operation of the active load balancer whenever a problem arises.
Our experts would be glad to discuss the selection and configuration of load balancers when developing an operation strategy for your specific application.
- Server load balancing
- Load balancing based on URLs
- Layer 7 healthchecks
- Load balancer failover