Exam DP-203 topic 2 question 51 discussion

B is correct answer. High concurrency cluster cannot terminated, so C is wrong. Standard cluster cannot shared by multiple tasks, so A and D are wrong.

The cluster does auto-scaling by default. Auto-termination should be set up manually

I think its C. Firstly I couldn't find anything related to autoscaling at cluster level . The autoscaling is enabled by default at the pipeline. So I don't find out anything that mention how to set up auto scaling for High concurrency cluster so ruling out B.

B is correct

Here is my take on this. Autoscaling addresses first requirement i-e latency and Auto terminate addresses third requirement i-e cost. Now we have to implement both of them manually, meaning none of them is by default. But in the question it says reduce cost without compromising other requirements. So if we go for auto-termination that means we are not autoscaling, compromising the first requirement i-e the latency requirement. therefore the given option is correct i-e B: High Concurrency with autoscaling

Agree to B

B is correct

Just because auto termination is eligible for high concurrency clusters, doesn't mean we have to use it. A key requirement is to "minimize query latency", which makes autoscaling more favorable. Ref: "Workloads can run faster compared to a constant-sized under-provisioned cluster." https://docs.microsoft.com/en-us/azure/databricks/clusters/configure#cluster-size-and-autoscaling

High Concurrency clusters can be configured with auto termination (I just checked). BUT: The questions says: reduce costs WITHOUT compromising the other requirements. So I would still go for autoscaling, since there is no answer option that offers both (autoscaling and auto termination)

Answer should be B as per below The key benefits of High Concurrency clusters are that they provide fine-grained sharing for maximum resource utilization and minimum query latencies.Autoscaling clusters can reduce overall costs compared to a statically-sized cluster.

i try it and it´s possible to create a cluster with auto termination.

https://docs.microsoft.com/en-us/azure/databricks/clusters/configure#cluster-mode - High Concurrency clusters do not terminate automatically by default. - A Standard cluster is recommended for a single user.

I would say C. High Concurrency with Auto Termination. Although the default is no auto terminate we can still overwrite that setting.

B is correct answer. High concurrency cluster cannot AUTO terminated

Auto terminate for high concurrency cluster is possible. But due to the 2nd point 'Maximize the number of users that can run queries on the cluster at the same time', I will go with option B. High Concurrency and Auto Scaling

it's correct