diff options
| author | Joshua Lambert <joshua@gitlab.com> | 2017-09-15 00:08:58 -0400 |
|---|---|---|
| committer | Joshua Lambert <joshua@gitlab.com> | 2017-09-15 00:08:58 -0400 |
| commit | 5bc44fed335bc04c2e0db033087560c2059ffb8c (patch) | |
| tree | fad2311cb2e6ef089090ea9f58b6d4a52ec81700 | |
| parent | ec3b3797e789d82200de2694458cc7cde2093549 (diff) | |
| download | gitlab-ce-37895-kubernetes-cpu-metrics-should-be-avg-not-sum.tar.gz | |
fix k8s cpu metrics)37895-kubernetes-cpu-metrics-should-be-avg-not-sum
| -rw-r--r-- | config/prometheus/additional_metrics.yml | 8 | ||||
| -rw-r--r-- | doc/user/project/integrations/prometheus_library/kubernetes.md | 2 |
2 files changed, 4 insertions, 6 deletions
diff --git a/config/prometheus/additional_metrics.yml b/config/prometheus/additional_metrics.yml index 33b897f46e2..f7b82f1cc22 100644 --- a/config/prometheus/additional_metrics.yml +++ b/config/prometheus/additional_metrics.yml @@ -149,13 +149,11 @@ label: Average unit: MB - title: "CPU Utilization" - y_label: "CPU Utilization (%)" + y_label: "CPU Utilization (Cores)" required_metrics: - container_cpu_usage_seconds_total weight: 1 queries: - - query_range: 'sum(rate(container_cpu_usage_seconds_total{container_name!="POD",%{environment_filter}}[2m])) by (cpu) * 100' + - query_range: 'sum(rate(container_cpu_usage_seconds_total{container_name!="POD",%{environment_filter}}[2m]))' label: CPU - unit: "%" - series: - - label: cpu + unit: "cores" diff --git a/doc/user/project/integrations/prometheus_library/kubernetes.md b/doc/user/project/integrations/prometheus_library/kubernetes.md index 4d39ae0c4fa..4e01d6aac2f 100644 --- a/doc/user/project/integrations/prometheus_library/kubernetes.md +++ b/doc/user/project/integrations/prometheus_library/kubernetes.md @@ -8,7 +8,7 @@ GitLab has support for automatically detecting and monitoring Kubernetes metrics | Name | Query | | ---- | ----- | | Average Memory Usage (MB) | (sum(container_memory_usage_bytes{container_name!="POD",%{environment_filter}}) / count(container_memory_usage_bytes{container_name!="POD",%{environment_filter}})) /1024/1024 | -| Average CPU Utilization (%) | sum(rate(container_cpu_usage_seconds_total{container_name!="POD",%{environment_filter}}[2m])) by (cpu) * 100 | +| Average CPU Utilization (%) | sum(rate(container_cpu_usage_seconds_total{container_name!="POD",%{environment_filter}}[2m])) | ## Configuring Prometheus to monitor for Kubernetes node metrics |