managed services in relation to 'reliable and highly available system'

[kung]

Question was “your role is to implement a reliable and highly available system that can recover from infrastructure or service disruptions…”

I picked “Use managed and application level services to reduce cost of ownership.”
but [above] “is incorrect because this is only concerned about costing and not for reliability nor for ensuring high availability.”

Irrespective of the ‘to reduce cost of ownership’, managed services will IMO improve reliability/availability, might dynamically scale and most likely will mitigate service disruptions.

I.e. I think picking managed services is always the smarter option (if it fits technically and financially).

Furthermore, the selected correct answer talks about EC2 instances, but maybe you want to go serverless in your architecture? There’s nothing about this aspect in the scenario.

So I find it a bit ‘to easy’ to discard “Use managed and application level services” as a correct answer…

Cheers,
Robert

[Jon-Bonso]

Hi Robert,

Thanks for sharing your thoughts on this item. You can use a serverless option here, the only thing missing is a Regional failover configuration. The option says:

Use serverless architecture which eliminates the need for you to run and maintain servers or carry out traditional compute activities.

The correct answer is:

Scale horizontally to increase aggregate system availability. Replace one large resource with multiple small resources to reduce the impact of a single failure on the overall system. Distribute requests across multiple, smaller resources to ensure that they don’t share a common point of failure. Implement Auto-Scaling on your EC2 instances and utilize multiple Availability Zones. Clone your stack to another AWS Region and implement a Route 53 failover routing policy.

Serverless architectures are not immune to misconfigurations and disruptions. Yes, AWS handles the server management of your computing capacity, but you still have to do certain steps in order to improve its availability and reliability. AWS Lambda and API Gateway are Regional services, thus, if the AWS Region goes down, then your serverless application will be unavailable.

Nonetheless, I agree that the option that says: “managed and application-level services” can be revised to avoid any issues. I have made the necessary change and it will be reflected in our practice tests soon.

Cheers,

Jon Bonso