5 Resilient Deployment Architecture (Non-normative)

This section addresses Req. D1.2-2.8-Avail.

For TAS³ services to be dependable, they need to be deployed so that they are resilient to system and network failure. Resiliency and efficiency are the first lines of defense against Denial of Service attacks that try to attack simple catastrophic vulnerabilities or overwhelm the system on the point where it is most inefficient. Resiliency needs to be considered at several layers, namely on the Front Channel and on the Back Channel.

Fig-10: Layering of resilience features for Front Channel, Back Channel, and data center Back End services.

Fig-11: Resiliency implemented using hardware load balancers.

Note that the virtual IP address is hosted either in hardware load balancer, or one member of a cluster. Fail-over of the virtual IP is arranged using Virtual Router Redundancy Protocol (VRRP) [RFC3768].

Fig-12: Resiliency implemented using software load-balancing-fail-over functionality and clustering.