Databricks Certified Data Engineer - Professional
Get started today
Ultimate access to all questions.
Comments
Loading comments...
Ultimate access to all questions.
In the context of optimizing a Spark application's performance using the Spark UI, Ganglia UI, and the Cluster UI, consider the following scenario: Your application is experiencing slow performance and occasional failures. You need to diagnose the issue and implement the most effective solution to improve performance. The solution must consider cost efficiency, scalability, and minimal downtime. Which of the following actions should you take based on the insights provided by these UIs? (Choose one correct option)
A
Ignore the Spark UI's indication of high garbage collection time, as it does not significantly impact application performance.
B
Upon noticing high CPU and memory usage on a specific node via the Ganglia UI, immediately conclude that scaling up the node is the only solution without considering other optimization techniques.
C
Use the Spark UI to identify long-running stages and tasks, then optimize the application by repartitioning the data or refining transformations to reduce execution time and resource consumption.
D
If the Cluster UI indicates a node failure, permanently remove the node from the cluster without attempting to recover or replace it, assuming it's a hardware failure.