Users report increased HTTP 5xx responses. Separate responses generated by the ALB from responses generated by targets.
Compare response counts for the same load balancer and incident interval in CloudWatch.
Which metrics and statistic are most appropriate?
Compare the same-period Sum of ELB and Target 5xx metrics to identify the response generator and guide investigation.
Detailed explanation
These compare 4xx rather than the required 5xx category.
These compare 4xx rather than the required 5xx category.
Target health counts do not identify the generator of a 5xx response.
Target health counts do not identify the generator of a 5xx response.
These count 5xx responses generated by the ALB and by targets separately.
These count 5xx responses generated by the ALB and by targets separately.
Latency and request volume do not separate 5xx counts by response generator.
Latency and request volume do not separate 5xx counts by response generator.
Check official documentation
Compare the following points with the official documentation.
Read both 5xx metric definitions, their Sum statistic, and access-log status-code fields.
Official references
What to confirm
Identify the response generator without assuming that an ELB-generated error excludes a target-side cause.
Key points
- HTTPCode_ELB_5XX_Count
- HTTPCode_Target_5XX_Count
- ALB
Notes
- Environment: AWS SOA-C03 exam guide and the linked AWS documentation
- Reviewed against the linked AWS documentation on 2026-10-03.
Foundation review
Align the aggregation
Use AWS/ApplicationELB metrics for the same LoadBalancer and time window with Sum. Average does not represent the total response count.
Generator versus root cause
An ALB-generated error can still involve target connectivity or response behavior. Compare elb_status_code and target_status_code in enabled access logs, application logs, and target health before attributing the cause.