Data Warehousing on AWS Training Data Warehousing on AWS: Workload Management 5 — Questions and Answers
Question 1: A Redshift cluster has 5 WLM queues configured with memory percentages of 20%, 20%, 20%, 20%, and 15%. What percentage is left for the default queue?
- 5% (Correct answer)
- 10%
- 15%
- 20%
Correct answer: 5%
Total WLM memory must equal 100%; with 95% allocated across five queues (20+20+20+20+15), only 5% remains for the default queue.
Question 2: Which feature allows Redshift to automatically scale read capacity to handle unpredictable spikes in concurrent users?
- Elastic resize
- Concurrency scaling (Correct answer)
- RA3 node auto-tiering
- Workload isolation groups
Correct answer: Concurrency scaling
Concurrency scaling automatically adds transient read capacity when query queuing occurs, enabling consistent performance during demand spikes without manual intervention.
Question 3: In Redshift WLM, what does the 'queue assignment rules' evaluation order determine?
- The sequence in which memory is allocated to queues at startup
- Which queue a query is routed to when multiple queues have matching criteria (Correct answer)
- The order in which queries within a single queue are executed
- The priority of vacuum operations relative to user queries
Correct answer: Which queue a query is routed to when multiple queues have matching criteria
Redshift evaluates WLM queue assignment rules top-to-bottom; a query is assigned to the first queue in configuration order whose user group or query group criteria match.
Question 4: What Redshift WLM configuration approach is recommended by AWS for most workloads as of recent best practices?
- Manual WLM with 8 queues and fixed memory percentages
- Automatic WLM with query priorities (Correct answer)
- Default WLM with no customization
- Separate clusters per workload type instead of WLM queues
Correct answer: Automatic WLM with query priorities
AWS recommends automatic WLM with priority settings because it uses machine learning to dynamically optimize concurrency and memory, reducing the need for manual tuning.
Question 5: Which of the following is a valid use case for setting a WLM query timeout in Redshift?
- Preventing runaway queries from consuming cluster resources indefinitely (Correct answer)
- Automatically compressing query results after a time threshold
- Scheduling queries to run at specific times of day
- Limiting the number of rows returned by long-running queries
Correct answer: Preventing runaway queries from consuming cluster resources indefinitely
WLM query timeouts cancel queries that exceed the configured duration, protecting the cluster from runaway queries that would otherwise monopolize resources.
Question 6: How can you monitor the amount of time queries spend waiting in a Redshift WLM queue over time?
- Query the SVL_WLM_QUERY_QUEUE_INFO view for queue wait durations (Correct answer)
- Check the Redshift console billing dashboard
- Review the S3 access logs for the associated data bucket
- Inspect the VPC Flow Logs for the cluster subnet
Correct answer: Query the SVL_WLM_QUERY_QUEUE_INFO view for queue wait durations
SVL_WLM_QUERY_QUEUE_INFO contains per-query wait time metrics showing how long each query spent waiting in its assigned WLM queue before execution began.
Question 7: When concurrency scaling is enabled, which queues are eligible to use the transient concurrency scaling clusters?
- Only the default WLM queue
- Any queue with the concurrency scaling mode set to 'auto' (Correct answer)
- Only queues with superuser access
- All queues regardless of configuration
Correct answer: Any queue with the concurrency scaling mode set to 'auto'
Concurrency scaling must be explicitly enabled per queue by setting the queue's concurrency scaling mode to 'auto'; queues left on 'off' never spill to scaling clusters.
Data Warehousing on AWS Training Data Warehousing on AWS: Workload Management 5
A Redshift cluster has 5 WLM queues configured with memory percentages of 20%, 20%, 20%, 20%, and 15%.
What percentage is left for the default queue?