Which metric is least useful for identifying ECS infrastructure issues?

The correct answer is B. Number of active user sessions. Number of active user sessions (B) is the least useful because it's an application-level business metric - it reflects user behavior, not the health of the underlying ECS infrastructure. The other three are genuine infrastructure signals: disk space availability (A) directly…

ECS Performance Design

Question

Options

ADisk space availability
BNumber of active user sessions
CCPU steal time
DMemory swap usage

How the community answered

(56 responses)

A
7% (4)
B
73% (41)
C
4% (2)
D
16% (9)

Explanation

Number of active user sessions (B) is the least useful because it's an application-level business metric - it reflects user behavior, not the health of the underlying ECS infrastructure. The other three are genuine infrastructure signals: disk space availability (A) directly impacts ECS's ability to pull container images and write logs, and exhaustion causes task failures; CPU steal time (C) reveals contention on the underlying hypervisor, meaning your EC2 host is starved of physical CPU cycles - a silent killer of container performance; memory swap usage (D) indicates the host is under memory pressure, which degrades container throughput and can trigger OOM kills. Memory tip: Think "infrastructure = resources the OS manages" - disk, CPU cycles, and RAM are OS-level; user sessions live above the infrastructure layer and belong in application monitoring, not ECS health dashboards.

Topics

#infrastructure monitoring#performance metrics#user sessions#resource utilization

Community Discussion

No community discussion yet for this question.

Full D-ECS-DS-23 Practice