You want to alert when the Kafka broker's RequestQueueSize exceeds 100 for more than 5 minutes. Which metric and approach should you use?

hard🚀 Application Q8 of Q15

Kafka - Monitoring and Operations

You want to alert when the Kafka broker's RequestQueueSize exceeds 100 for more than 5 minutes. Which metric and approach should you use?

Monitor <code>kafka.network:type=RequestMetrics,name=QueueSize,request=Produce</code> and trigger alert if value > 100 for 5 minutes

BMonitor <code>kafka.network:type=RequestMetrics,name=RequestQueueSize</code> without request type and alert immediately

CUse <code>UnderReplicatedPartitions</code> metric to detect queue size issues

DMonitor <code>RequestHandlerAvgIdlePercent</code> and alert if below 50%

Step-by-Step Solution

Solution:

Step 1: Identify the correct metric for request queue size
The metric with name=QueueSize and request=Produce is the correct one to monitor queue size for produce requests.
Step 2: Define alert condition
Trigger alert if the metric value exceeds 100 continuously for 5 minutes to avoid false positives.
Final Answer:
Monitor kafka.network:type=RequestMetrics,name=QueueSize,request=Produce and trigger alert if value > 100 for 5 minutes -> Option A
Quick Check:
QueueSize metric + threshold + duration = Alert condition [OK]

Quick Trick: Use specific request type in QueueSize metric for accurate alerts [OK]

Common Mistakes:

MISTAKES

Master "Monitoring and Operations" in Kafka

9 interactive learning modes - each teaches the same concept differently

Want More Practice?

15+ quiz questions · All difficulty levels · Free

More Kafka Quizzes