1Z0-1067-25 Question 44
Select 2Your team manages a data transformation pipeline in Oracle Cloud Infrastructure that processes large datasets daily. The pipeline sees a four-hour peak usage window for data ingestion and transformation every morning, with minimal processing the rest of the day. You need to minimize costs while ensuring your pipeline has sufficient performance during peak times. Which two strategies should you implement?
- A
Keep a single large compute shape running continuously to handle all loads
- B
Use an instance pool with autoscaling enabled based on performance metrics
- C
Switch to Exadata shapes to guarantee maximum performance
- D
Leverage flexible compute shapes with CPU and memory scaling policies
- E
Over-provision compute resources to avoid under-capacity
Show answer and explanation
Correct answers: B, D
Explanation
To optimize cost and performance for variable workloads on Oracle Cloud Infrastructure, using instance pools with autoscaling and flexible shapes is a proven best practice. Autoscaling automatically adjusts instance counts based on demand to handle peak loads efficiently and scale down during idle times. Similarly, flexible shapes let you dynamically allocate CPU and memory resources, which reduces over-provisioning and avoids paying for idle capacity. Refer to Oracle� documentation on autoscaling (docs.oracle.com/en/oci) and resource-based scaling to ensure proper configurations that balance cost and performance requirements.
- A. Incorrect.
Option 1: Incorrect. Keeping a single large shape running 24/7 is not cost-efficient. Even though it meets peak demand, you pay for idle resources during off-peak hours.
- B. Correct.
Option 2: Correct. Instance pools with autoscaling dynamically add or remove instances based on usage. This strategy ensures plenty of capacity at peak times and reduced cost during low usage.
- C. Incorrect.
Option 3: Incorrect. Exadata shapes are specialized for Oracle Database workloads and can be costly. They are not necessarily the best fit for a general data transformation pipeline unless your workload specifically requires Exadata features.
- D. Correct.
Option 4: Correct. Flexible compute shapes allow you to adjust CPU and memory resources automatically. Scaling these resources during peak hours and reducing them afterward minimizes costs while delivering required performance.
- E. Incorrect.
Option 5: Incorrect. Over-provisioning guarantees performance but leads to unnecessary costs. The objective is to match capacity to workload demands rather than always running at full capacity.