3.3 Load Balancing

Question 1

Explain what Load Balancing is.

Answer 1

A network load balancer (NLB) is a device that is used to direct traffic to an array of web servers, applications servers, or other service endpoints.

NLB = Network load balancer = Load balancer = LB

Question 2

Explain Load balancer Configurations.

Answer 2

-There are several ways to setup a load balancer (LB).

–Active/Active - the load balancers act like an array, dealing with the traffic together as both are active.
–Single LB failure may degrade peformance.

-Active/Passive - The active node is fulfiling load balancing duties and the passive node is listening and monitoring the active node.
Should be active node fail, then the passive node will take over, providing redundancy.

Question 3

Explain a Virtual IP

Answer 3

A virtual IP address eliminates a host’s dependency upon individual network interfaces.

Web traffic comes into the NLB from the virtual IP address (VIP) on the frontend.

Request is sent to one of the web servers in the server farm (on the backend).

Question 4

Explain Load Balance Scheduling.

Answer 4

Scheduling options, which determine how the load is distributed by the load balancer, inlclude:

-Lease Utilized Host: NLB knows the status of all servers in the server farms and which web servers are the least utilized by using a scheduling algorithm.

-DNS Round Robin: When the request comes in, the load balancer contacts the DNS server and rotates the request based on the lowest IP address first.

-Affinity: When the LB is set to Affinity, the request is sent to the same web server based on the requester’s IP address, IP+port, and/or session ID
*Affinity configuration may be referred to in tuples (2-tuple, 3-tuple)
This is also known as persistence or a sticky session, where the load balancer uses the same server for the session.