Why Study with PlanetCert?
The Latest Questions
Practice questions and exam topics aligned with the current exam objectives.
Detailed Explanations
Go beyond the answer. Master the material with comprehensive learning and professional explanations for every concept.

AI-Powered Insights
Personalized preparation guidance that adapts to your performance and identifies weak spots automatically.
Exam Information
Official specifications published by Databricks
Exam Format
Registration
Validity
Skills this practice test covers
- Explain lazy evaluation and how an action becomes jobs, stages and tasks.
- Read and write DataFrames with DataFrameReader and DataFrameWriter, schemas and save modes.
- Join, union, filter and aggregate DataFrames, and explode arrays into rows.
- Choose between repartition and coalesce, and find data skew and needless shuffles.
- Write Streaming DataFrames with the right output mode and sink, and drop duplicates with a watermark.
- Connect to a cluster through Spark Connect and run pandas code with the Pandas API on Spark.
Explore related topics
Apache-Spark-Developer exam content
7 domains · 35 topics · 192 of 210 questions sorted by domain
Apache Spark Architecture and Components20%34 questions
This domain explains how Spark runs code on a cluster. Candidates describe the driver, executors, cores and memory, the execution hierarchy of jobs, stages and tasks, and how lazy evaluation holds back transformations until an action runs.
Skills tested
- Describe the roles of the driver, the executors and the cluster manager.
- Explain how an action triggers a job that splits into stages and tasks.
- Describe the SparkSession lifecycle, caching and storage levels.
- Configure partitioning for distributed data processing.
- Identify the features of Spark modules such as Spark SQL, Structured Streaming and MLlib.
Questions by topic
- Explain the Apache Spark Architecture execution hierarchy.6
- Identify the role of core components of Apache Spark's Architecture including cluster, driver node, worker nodes/executors, CPU cores, memory6
- Describe the execution patterns of the Apache Spark engine, including actions, transformations, and lazy evaluation10
- Describe the architecture of Apache Spark, including DataFrame and Dataset concepts, SparkSession lifecycle, caching, storage levels, and garbage collection6
- Describe the different deployment mode types (Client, Cluster, Local) in Apache Spark environment2
- Describe the architecture of Apache Spark, including DataFrame and Dataset concepts4
Using Spark SQL20%26 questions
This domain is about data sources, files and tables. Candidates read and write data through JDBC and files, run SQL directly on files, save persistent tables with sorting and partitioning, and register DataFrames as temporary views.
Skills tested
- Read from and write to JDBC sources and files, including overwrites and partitioning by column.
- Run SQL queries directly on ORC, JSON, CSV, text and Delta files.
- Save data to a persistent table with sorting and partitioning for faster reads.
- Register a DataFrame as a temporary view and query it with SQL.
Questions by topic
- Execute SQL queries directly on files including ORC Files, JSON Files, CSV Files, Text Files, and Delta Files8
- Utilize common data sources such as JDBC, files, etc. to efficiently read from and write to Spark DataFrames using SparkSQL, including overwriting and partitioning by column7
- Register DataFrames as temporary views in Spark SQL, allowing them to be queried with SQL syntax.5
- Utilize common data sources such as JDBC, files, etc. to efficiently read from and write to Spark DataFrames using SparkSQL5
- Save data to persistent tables while applying sorting, and partitioning to optimize data retrieval1
Developing Apache Spark DataFrame/DataSet API Applications30%62 questions
The largest domain is everyday DataFrame code. Candidates change columns and rows, remove duplicates, aggregate, work with dates, combine DataFrames with joins and unions, and write user-defined functions. Broadcast variables, accumulators and broadcast joins belong here too.
Skills tested
- Add, drop, rename and split columns, filter rows and explode arrays.
- Aggregate with count, approximate count distinct, mean and summary.
- Convert a Unix epoch to a date string and extract date components.
- Combine DataFrames with inner, left, cross and broadcast joins, and with union.
- Create and invoke user-defined functions, and explain broadcast variables and accumulators.
Questions by topic
- Describe different types of variables in Spark including broadcast variables and accumulators7
- Combine DataFrames with operations such as Inner join, left join, broadcast join, multiple keys, cross join, union, union all8
- Describe the purpose and implementation of broadcast joins6
- Manipulate columns, rows, and table structures by adding, dropping, splitting, renaming column names, applying filters, and exploding arrays14
- Perform data deduplication and validation operations on DataFrames5
- Manipulate and utilize Date data type such as Unix epoch to date string, extract date component5
- Manage input and output operations by writing, overwriting, and reading DataFrames with schemas4
- Perform aggregate operations on DataFrames such as count, approximate count distinct, and mean, summary9
- Create and invoke user-defined functions with or without stateful operators including StateStores3
- Utilize common data sources such as JDBC, files, etc. to efficiently read from and write to Spark DataFrames using SparkSQL, including overwriting and partitioning by column1
Troubleshooting and Tuning Apache Spark DataFrame API Applications10%26 questions
This domain starts from a slow or failing job. Candidates tune partitioning with repartition and coalesce, find data skew, reduce shuffling and explain Adaptive Query Execution (AQE). They also read driver and executor logs to diagnose out-of-memory errors and underused clusters.
Skills tested
- Pick repartition or coalesce to set the number of partitions.
- Identify data skew and reduce shuffling in a slow job.
- Explain what Adaptive Query Execution changes while a query runs.
- Read driver and executor logs to find out-of-memory errors and cluster underutilization.
Questions by topic
- Implement performance tuning strategies & optimize cluster utilization including identifying data skew3
- Implement performance tuning strategies & optimize cluster utilization including partitioning, repartitioning, coalescing6
- Describe Adaptive Query Execution (AQE) and its benefits.5
- Implement performance tuning strategies & optimize cluster utilization including partitioning, repartitioning, coalescing, identifying data skew, and reducing shuffling7
- Perform logging and monitoring of Spark applications – publish, customize, and analyze Driver logs and Executor logs to diagnose out-of-memory errors, cluster underutilization, etc.5
Structured Streaming10%20 questions
Structured Streaming treats a stream as a table that keeps growing. Candidates explain the micro-batch engine and its fault tolerance, create and write Streaming DataFrames with output modes and sinks, and run selections, windows, aggregations and deduplication with and without a watermark.
Skills tested
- Create a Streaming DataFrame with DataStreamReader and write it with DataStreamWriter.
- Choose the output mode and output sink for a streaming query.
- Apply windows and aggregations to a stream.
- Remove duplicate events with and without a watermark.
- Explain exactly-once semantics and checkpointing.
Questions by topic
- Create and write Streaming DataFrames and Streaming Datasets including the basic output modes and output sinks8
- Perform Streaming Deduplication in Structured Streaming, both with and without watermark usage4
- Explain the Structured Streaming engine in Spark, including its functions, programming model, micro-batch processing3
- Perform basic operations on Streaming DataFrames and Streaming Datasets such as selection, projection, window and aggregation4
- Explain the Structured Streaming engine in Spark, including its functions, programming model, micro-batch processing, exactly-once semantics, and fault tolerance mechanisms1
Using Spark Connect to deploy applications5%14 questions
Spark Connect separates the client application from the Spark cluster that does the work. Candidates describe its features and compare the client, cluster and local deployment modes.
Skills tested
- Describe how a Spark Connect client sends DataFrame operations to a remote Spark server.
- Compare client, cluster and local deployment modes.
Questions by topic
- Describe the features of Spark Connect7
- Describe the different deployment mode types (Client, Cluster, Local) in Apache Spark environment7
Using Pandas API on Apache Spark5%10 questions
The Pandas API on Spark runs familiar pandas code on a cluster. Candidates explain its advantages and create and invoke Pandas UDFs.
Skills tested
- Explain when the Pandas API on Spark helps a team that already uses pandas.
- Create and invoke a Pandas UDF.
Questions by topic
- Create and invoke Pandas UDF4
- Explain the advantages of using Pandas API on Spark6
Requirements
- No formal prerequisites: anyone can book the exam.
- Recommended: at least six months of practical Apache Spark development.
- Recommended: working Python skills and an understanding of distributed computing.
About the Apache-Spark-Developer exam
The Apache-Spark-Developer exam is the Databricks associate exam for developers who write Apache Spark applications in Python. Its largest domain is the DataFrame API: selecting, filtering, joining and aggregating data, handling dates and writing user-defined functions. The other domains cover how Spark runs a job on a cluster, Spark SQL, tuning, Structured Streaming, Spark Connect and the Pandas API on Spark. Passing earns the Databricks Certified Associate Developer for Apache Spark certification, which stays valid for two years.
The DataFrame API gets the most questions in the PlanetCert practice test, with joins, aggregations, date functions and user-defined functions (UDFs). Close pairs, such as repartition and coalesce or the append and complete output modes, are where the explanations matter most. Study Mode suits the tuning questions, where the cause of a slow job has to be found before the fix. Brainy AI Assistant can explain why a step causes a shuffle or rewrite a row-by-row Python UDF as a Pandas UDF.
How to earn the certification
- Earns
- Databricks Certified Associate Developer for Apache SparkAssociate
The certification stays valid for two years.
- After
- DATA-ENG-PRO Data Engineer Professional
The professional exam for engineers who build and run production data pipelines on Databricks.
- ML-PRO Machine Learning Professional
The professional exam for those who move from data work to building and running machine learning models.
- DATA-ENG-PRO Data Engineer Professional
How to prepare for Apache-Spark-Developer
- 1
Run PySpark on a laptop
Install PySpark locally, or use any Spark cluster, and load a CSV file and a JSON file into DataFrames. Typing the methods by hand makes the API questions easier to read.
- 2
Watch a job in the Spark UI
Run a join and a groupBy, then open the Spark UI to see the jobs, stages and tasks they create. This ties the architecture domain to real code.
- 3
Drill the DataFrame API daily
The DataFrame/DataSet API domain is the largest. Work through column changes, joins, aggregations, date functions and UDFs until each method's arguments come without a lookup.
- 4
Break a streaming query on purpose
Start a Structured Streaming query on a folder of files, switch between output modes, and add a watermark to a deduplication. Note which combinations Spark rejects and why.
- 5
Sit one timed practice run
Take the Apache-Spark-Developer practice test in Timed Mode. For each missed tuning or streaming question, ask Brainy AI Assistant to explain what Spark does at run time.
Start a timed Apache-Spark-Developerrun
Who Apache-Spark-Developer is for
- ETL developers and data engineers who build Spark pipelines in Python.
- Software engineers who move into big data work.
- Analytics engineers who move from SQL alone to the DataFrame API.
Typical job titles: Spark developer · Data engineer · Big data developer · ETL developer · Analytics engineer
Apache-Spark-Developer glossary
17 terms from the exam objectives and the vendor's official material, each defined in a sentence.
- Accumulators
- Shared variables that tasks can only add to, such as a counter, and that the driver reads.
- Adaptive Query Execution (AQE)
- A Spark SQL feature that re-plans a query while it runs, using real statistics, for example to coalesce shuffle partitions or to split skewed joins.
- Broadcast join
- A join in which Spark copies a small table to every executor, so the large table does not need a shuffle.
- Broadcast variables
- Read-only values that Spark sends once to each executor instead of with every task.
- Catalyst Optimizer
- The Spark SQL query optimizer, which turns DataFrame and SQL code into an efficient physical plan.
- Coalescing
- Reducing the number of partitions without a full shuffle, which is cheaper than repartitioning.
- Data skew
- A few partitions hold much more data than the others, so their tasks run far longer than the rest.
- Exactly-once semantics
- Each record affects the result exactly once, even after a failure and a restart.
- Jobs, stages, and tasks
- The execution hierarchy: an action starts a job, the job splits into stages at shuffle boundaries, and each stage runs as parallel tasks.
- Lazy evaluation
- Spark records transformations without running them, and runs them only when an action asks for a result.
- Micro-batch processing
- The Structured Streaming model that processes new stream data in a series of small batches.
- Output modes
- The rule for what a streaming query writes on each trigger: append, complete or update.
- Pandas UDF
- A user-defined function that Spark runs on whole batches of data with pandas, which is faster than a row-by-row Python UDF.
- Repartitioning
- Changing the number of partitions with a full shuffle, so that data is spread evenly again.
- Save modes
- The settings that decide what happens when output data already exists: append, overwrite, ignore or error.
- Spark Connect
- A client-server interface that lets an application send DataFrame operations to a remote Spark cluster.
- Watermark
- A limit on how late events may arrive, which lets Spark drop old state during aggregation or deduplication.