Elastic Load Balancer with Duration-Based Sticky Sessions

Answer - B and C.

When you create an elastic load balancer, a default level of capacity is allocated and configured.

As Elastic Load Balancing sees changes in the traffic profile, it will scale up or down.

The time required for Elastic Load Balancing to scale can range from 1 to 7 minutes, depending on the changes in the traffic profile.

When Elastic Load Balancing scales, it updates the DNS record with the new list of IP addresses.

To ensure that clients are taking advantage of the increased capacity, Elastic Load Balancing uses a TTL setting on the DNS record of 60 seconds.

It is critical that you factor this changing DNS record into your tests.

If you do not ensure that DNS is re-resolved or use multiple test clients to simulate increased load, the test may continue to hit a single IP address when Elastic Load Balancing has actually allocated many more IP addresses.

Because your end-users will not all be resolving to that single IP address, your test will not be a realistic sampling of real-world behavior.

Option A is incorrect because creating a load tester in US-East-1 is inconsequential since it is in a different region.

Option B is CORRECT because if you do not ensure that DNS is re-resolved, the test may continue to hit the single IP address.

Option C is CORRECT because if the requests come from globally distributed users, the DNS will not be resolved to a single IP address.

The traffic would be distributed evenly across multiple instances.

Option D is incorrect because the traffic will be routed to the same back-end instances as the users continue to access your application.

The load will not be evenly distributed across the AZs.

Please refer to the below article for more information.

http://aws.amazon.com/articles/1636185810492479

The scenario described above involves an Elastic Load Balancer with Duration-Based sticky sessions enabled in front of six EC2 web application instances spread across two Availability Zones. A software-based load tester is used to simulate traffic to the Elastic Load Balancer and several hundred users browse to the ELB's hostname.

The objective of load testing is to identify potential bottlenecks in the application or infrastructure under test, determine the maximum capacity that can be handled by the system, and measure the response time and performance metrics under different traffic scenarios.

Option A suggests creating a software-based load tester in US-East-1 and testing from there. However, this option does not address the high availability aspect of the architecture and may not accurately reflect the real-world scenario where users may be accessing the application from different locations.

Option B suggests forcing the software-based load tester to re-resolve DNS before every request. This option can help distribute the load across multiple EC2 instances behind the load balancer, but it does not address the need for global testing and may not accurately reflect the performance in real-world scenarios.

Option C suggests using a third-party load-testing service to send requests from globally distributed clients. This option can help simulate real-world traffic patterns, identify regional performance issues, and provide detailed performance metrics. However, this option may incur additional costs and may require additional setup and configuration.

Option D suggests switching to application-controlled sticky sessions. This option can help distribute the load across multiple EC2 instances based on the application's logic, but it may not be suitable for all applications and may require additional development effort.

In summary, Option C is the most suitable answer as it can provide a more realistic load testing scenario, identify potential performance issues, and provide detailed performance metrics from globally distributed clients. However, depending on the specific requirements and constraints of the project, other options may also be considered.