DEA-C01 Data Operations and Support Practice Question
A data engineer is monitoring an Amazon Kinesis Data Analytics for Apache Flink application that processes streaming data. The application is falling behind (increasing 'MillisBehindLatest') and the CPU utilization of the Flink task managers is consistently above 80%. Which THREE actions should the engineer take to improve performance? (Choose THREE.)
⚠ Common exam trap
Test-takers frequently confuse decreasing checkpoint intervals with improving performance, not realizing that more frequent checkpoints increase CPU and I/O overhead, making the lag worse.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Increase the number of shards in the Kinesis data stream.
Option A is correct because increasing the number of shards in the Kinesis data stream raises the stream's read throughput capacity (each shard supports up to 1 MB/s or 1,000 records/s for reads), which directly addresses the growing MillisBehindLatest by allowing the Flink source to consume data faster. Option C is correct because enabling auto-scaling for the Flink application (via Kinesis Data Analytics' automatic scaling of KPUs) dynamically adds parallel task slots and CPU resources when utilization is high, relieving the sustained >80% CPU pressure on task managers. Option E is correct because increasing the Flink application's parallelism distributes the workload across more parallel subtasks and task slots, lowering per-task CPU load and improving overall processing throughput. Option B is not appropriate because shortening the checkpoint interval increases checkpointing overhead and I/O rather than reducing state size, potentially worsening performance. Option D is wrong because decreasing the number of task managers reduces available CPU and parallelism, which would increase contention and further degrade the application.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Increase the number of shards in the Kinesis data stream.
Why this is correct
Adding shards raises the stream's ingest and read parallelism, letting more Flink subtasks consume concurrently and reducing the backlog driving MillisBehindLatest. However, shards only relieve source-side throughput; the sustained CPU above 80% on task managers still requires scaling parallelism or KPUs, so this alone is insufficient.
- ✗
Decrease the checkpoint interval to reduce state size.
Why it's wrong here
More frequent checkpoints add barrier alignment and snapshot overhead, consuming CPU that is already above 80%, so lag worsens. Shortening the interval is correct when recovery-time objectives demand faster restore points and CPU headroom exists, not under sustained saturation.
- ✓
Enable auto-scaling for the Flink application.
Why this is correct
Auto-scaling adds task manager parallelism when CPU stays above 80%, directly relieving the bottleneck causing rising MillisBehindLatest. Kinesis Data Analytics for Apache Flink adjusts parallelism automatically, so throughput scales without manual redeployment, satisfying the stem's CPU-saturation and lag constraints.
- ✗
Decrease the number of task managers to reduce CPU contention.
Why it's wrong here
Removing task managers cuts the parallel task slots available, lowering aggregate throughput and increasing backlog. Reducing task managers suits cost trimming during low-traffic periods; relieving CPU contention at 80% utilisation instead requires adding task managers or raising parallelism.
- ✓
Increase the Flink application's parallelism.
Why this is correct
Increasing parallelism distributes the workload across more task slots, directly relieving the sustained CPU saturation above 80% that is throttling throughput. With additional parallel subtasks, each processes a smaller share of the Kinesis shards, allowing the application to consume records faster and reduce the growing MillisBehindLatest lag.
Go deeper
Related to this question
About these practice questions
One of 1,321 original DEA-C01 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This DEA-C01 practice question is part of Courseiva's free Amazon Web Services certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the DEA-C01 exam.