Databricks Machine Learning Professional exam dumps

Databricks Machine Learning Professional practice question 189 of 280

Databricks Certified Machine Learning Professional. Professional level, Databricks. Free question with the correct answer and a full explanation.

Databricks Machine Learning Professional Question 189

Select 3

You have developed a batch inference pipeline using Databricks that processes a large dataset stored in a Delta table and outputs predictions back into another Delta table. Your team now requires the pipeline to handle real-time streaming data from a Kafka source instead. Which of the following steps are necessary to convert the batch pipeline into a streaming pipeline in Databricks?

  1. A

    Replace the batch read operation with a streaming read from Kafka using the readStream method.

  2. B

    Modify the model's inference logic to handle one record at a time instead of batch processing.

  3. C

    Replace the batch write operation with a streaming write using the writeStream method.

  4. D

    Enable time windowing in the streaming pipeline to handle data arriving late.

  5. E

    Use a trigger interval to control the frequency of processing in the streaming pipeline.

Show answer and explanation

Correct answers: A, C, E

Explanation

To convert a batch deployment pipeline to a streaming deployment pipeline in Databricks, you need to replace batch operations with their streaming equivalents, such as readStream and writeStream, and configure a trigger interval to define the processing frequency. Modifying the model's inference logic or enabling time windowing is not always required unless the use case specifically demands it.

  • A. Correct.

    Correct: To handle real-time data, you need to use a streaming source like Kafka and replace the batch read operation with a streaming readStream operation.

  • B. Incorrect.

    Incorrect: The model's inference logic can often remain unchanged as Spark processes micro-batches in streaming mode, so there is no requirement to modify it to handle one record at a time.

  • C. Correct.

    Correct: In a streaming pipeline, you must replace batch write operations with a streaming write operation using the writeStream method.

  • D. Incorrect.

    Incorrect: While time windowing is useful for certain use cases like aggregating data, it is not a mandatory step for converting a batch pipeline to a streaming pipeline.

  • E. Correct.

    Correct: Using a trigger interval is an important step to control how frequently the streaming pipeline processes incoming data.

Timed practice exam

Take a Databricks Machine Learning Professional practice test under exam conditions

60 questions in 120 minutes, drawn from this bank, with a score report and a per-question review when you finish.

Start timed exam

More Databricks Machine Learning Professional practice questions

  1. 190.You are tasked with converting an existing machine learning batch inference pipeline to a streaming...
  2. 191.You are tasked with converting a batch inference pipeline into a streaming inference pipeline in...
  3. 192.You are tasked with converting a batch machine learning inference pipeline into a streaming...
  4. 193.You are tasked with converting a batch machine learning deployment pipeline to a streaming...
  5. 194.You are tasked with deploying a Databricks machine learning model for real-time predictions. The...
Back to the Databricks Machine Learning Professional exam dumps