为什么 CPU 使用率高
“为什么 CPU 使用率高”是一种反复出现的性能模式,综合自 1 条已审核的发现。当前集群置信度较低。
Common symptoms
- 容器在 60 秒内被限流 1250 次(docker_stats: cpu_throttle_periods = 1250)
- cgroup v2 显示 850 个被限流周期(cgroup: nr_throttled = 850)
- 容器进程 CPU 占用率固定在 100%(top: cpu_percent = 100%)
Likely causes
- 容器限流(2 次)
- 一般性能异常(1 次)
Evidence to look for
docker_stats: cpu_throttle_periods = 1250 — 容器在 60 秒内被限流 1250 次
cgroup: nr_throttled = 850 — cgroup v2 显示 850 个被限流周期
top: cpu_percent = 100% — 容器进程 CPU 占用率固定在 100%
What it usually means
当 CPU 才是真正的瓶颈时,通常的解决方法是对热点路径进行 profiling,并减少计算工作量。在未先进行 profiling 的情况下直接水平或垂直扩容是低效的。
What to do next
- 1在容器资源限制中提高 CPU quota(低投入)— 当前限制对该工作负载来说过低
- 2对容器工作负载进行 profiling 以优化(中等投入)— 工作负载可能存在低效的 CPU 使用模式
Cluster analysis
low confidence1 findingEvidence patterns
Container throttling2x observed
- Container throttled 1250 times in 60s (cpu_throttle_periods: 1250)
- cgroup v2 shows 850 throttled periods (nr_throttled: 850)
General performance anomaly1x observed
- Container process pinned at 100% CPU (cpu_percent: 100%)
Most common fixes
- 1increase cpu quota in container resource limits1x recommendedlow effort
- 2profile container workload for optimization1x recommendedmedium effort
See what's causing your CPU
cpum.ai turns CPU, process, disk, and memory signals into plain-English explanations with evidence.
Open cpum.ai