Skip to content

feat: new container cpu usage recording rule using rate() #1025

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 11 commits into from
Mar 27, 2025
Merged

feat: new container cpu usage recording rule using rate() #1025

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
943b74d
chore: revert irate function back to rate
sleepyfoodie Feb 6, 2025
d9264b5
chore: keep irate for node_namespace_pod_ontainer and add rate in as …
sleepyfoodie Feb 7, 2025
c7ba140
chore: add test for apps
sleepyfoodie Feb 7, 2025
ae2c9a6
Update tests/tests.yaml
sleepyfoodie Feb 14, 2025
c2a9315
chore: add 5m to queries to match recording name
sleepyfoodie Feb 14, 2025
13c42ae
chore: update readme
sleepyfoodie Feb 14, 2025
b39c5e8
update readme
sleepyfoodie Feb 14, 2025
48b28f4
chore: fix readme formatting
sleepyfoodie Feb 14, 2025
ff2894b
Update README.md
sleepyfoodie Feb 18, 2025
b94046a
Update dashboards/resources/node.libsonnet
sleepyfoodie Feb 18, 2025
c7cf95c
try adding space
sleepyfoodie Feb 18, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions dashboards/resources/cluster.libsonnet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -129,7 +129,7 @@ local var = g.dashboard.variable;
+ tsPanel.queryOptions.withTargets([
prometheus.new(
'${datasource}',
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The recording rule says node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate5m, but all the code changes say node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate.

)
+ prometheus.withLegendFormat('__auto'),
]),
Expand All @@ -144,23 +144,23 @@ local var = g.dashboard.variable;
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),

prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),

prometheus.new('${datasource}', 'sum(namespace_cpu:kube_pod_container_resource_requests:sum{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),

prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster"}) by (namespace) / sum(namespace_cpu:kube_pod_container_resource_requests:sum{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster"}) by (namespace) / sum(namespace_cpu:kube_pod_container_resource_requests:sum{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),

prometheus.new('${datasource}', 'sum(namespace_cpu:kube_pod_container_resource_limits:sum{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),

prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster"}) by (namespace) / sum(namespace_cpu:kube_pod_container_resource_limits:sum{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster"}) by (namespace) / sum(namespace_cpu:kube_pod_container_resource_limits:sum{%(clusterLabel)s="$cluster"}) by (namespace)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
])
Expand Down
8 changes: 4 additions & 4 deletions dashboards/resources/multi-cluster.libsonnet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,27 +99,27 @@ local var = g.dashboard.variable;
cpuUsage: [
tsPanel.new('CPU Usage')
+ tsPanel.queryOptions.withTargets([
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate) by (%(clusterLabel)s)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate) by (%(clusterLabel)s)' % $._config)
+ prometheus.withLegendFormat('__auto'),
]),
],

cpuQuota: [
g.panel.table.new('CPU Quota')
+ g.panel.table.queryOptions.withTargets([
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate) by (%(clusterLabel)s)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate) by (%(clusterLabel)s)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(kube_pod_container_resource_requests{%(kubeStateMetricsSelector)s, resource="cpu"}) by (%(clusterLabel)s)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate) by (%(clusterLabel)s) / sum(kube_pod_container_resource_requests{%(kubeStateMetricsSelector)s, resource="cpu"}) by (%(clusterLabel)s)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate) by (%(clusterLabel)s) / sum(kube_pod_container_resource_requests{%(kubeStateMetricsSelector)s, resource="cpu"}) by (%(clusterLabel)s)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(kube_pod_container_resource_limits{%(kubeStateMetricsSelector)s, resource="cpu"}) by (%(clusterLabel)s)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate) by (%(clusterLabel)s) / sum(kube_pod_container_resource_limits{%(kubeStateMetricsSelector)s, resource="cpu"}) by (%(clusterLabel)s)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate) by (%(clusterLabel)s) / sum(kube_pod_container_resource_limits{%(kubeStateMetricsSelector)s, resource="cpu"}) by (%(clusterLabel)s)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
])
Expand Down
12 changes: 6 additions & 6 deletions dashboards/resources/namespace.libsonnet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,15 +91,15 @@ local var = g.dashboard.variable;
statPanel(
'CPU Utilisation (from requests)',
'percentunit',
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"}) / sum(kube_pod_container_resource_requests{%(kubeStateMetricsSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace", resource="cpu"})' % $._config
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace"}) / sum(kube_pod_container_resource_requests{%(kubeStateMetricsSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace", resource="cpu"})' % $._config
)
+ stat.gridPos.withW(6)
+ stat.gridPos.withH(3),

statPanel(
'CPU Utilisation (from limits)',
'percentunit',
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"}) / sum(kube_pod_container_resource_limits{%(kubeStateMetricsSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace", resource="cpu"})' % $._config
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace"}) / sum(kube_pod_container_resource_limits{%(kubeStateMetricsSelector)s, %(clusterLabel)s="$cluster", namespace="$namespace", resource="cpu"})' % $._config
)
+ stat.gridPos.withW(6)
+ stat.gridPos.withH(3),
Expand All @@ -125,7 +125,7 @@ local var = g.dashboard.variable;
+ tsPanel.queryOptions.withTargets([
prometheus.new(
'${datasource}',
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config
)
+ prometheus.withLegendFormat('__auto'),

Expand Down Expand Up @@ -197,19 +197,19 @@ local var = g.dashboard.variable;
table.new('CPU Quota')
+ table.gridPos.withW(24)
+ table.queryOptions.withTargets([
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", namespace="$namespace"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
])
Expand Down
8 changes: 4 additions & 4 deletions dashboards/resources/node.libsonnet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,7 +87,7 @@ local var = g.dashboard.variable;

prometheus.new(
'${datasource}',
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config,
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config,
)
+ prometheus.withLegendFormat('{{pod}}'),
])
Expand All @@ -105,19 +105,19 @@ local var = g.dashboard.variable;

table.new('CPU Quota')
+ table.queryOptions.withTargets([
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", node=~"$node"}) by (pod)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
])
Expand Down
8 changes: 4 additions & 4 deletions dashboards/resources/pod.libsonnet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,7 +94,7 @@ local var = g.dashboard.variable;
+ tsPanel.queryOptions.withTargets([
prometheus.new(
'${datasource}',
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{namespace="$namespace", pod="$pod", %(clusterLabel)s="$cluster", container!=""}) by (container)' % $._config
'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{namespace="$namespace", pod="$pod", %(clusterLabel)s="$cluster", container!=""}) by (container)' % $._config
)
+ prometheus.withLegendFormat('__auto'),

Expand Down Expand Up @@ -208,19 +208,19 @@ local var = g.dashboard.variable;
table.new('CPU Quota')
+ table.gridPos.withW(24)
+ table.queryOptions.withTargets([
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_requests{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container)' % $._config)
prometheus.new('${datasource}', 'sum(node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container) / sum(cluster:namespace:pod_cpu:active:kube_pod_container_resource_limits{%(clusterLabel)s="$cluster", namespace="$namespace", pod="$pod", container!=""}) by (container)' % $._config)
+ prometheus.withInstant(true)
+ prometheus.withFormat('table'),
])
Expand Down
2 changes: 1 addition & 1 deletion dashboards/resources/workload-namespace.libsonnet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,7 +91,7 @@ local var = g.dashboard.variable;

local cpuUsageQuery = |||
sum(
node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"}
node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace"}
* on(namespace,pod)
group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload_type=~"$type"}
) by (workload, workload_type)
Expand Down
2 changes: 1 addition & 1 deletion dashboards/resources/workload.libsonnet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -103,7 +103,7 @@ local var = g.dashboard.variable;

local cpuUsageQuery = |||
sum(
node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate{%(clusterLabel)s="$cluster", namespace="$namespace"}
node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate{%(clusterLabel)s="$cluster", namespace="$namespace"}
* on(namespace,pod)
group_left(workload, workload_type) namespace_workload_pod:kube_pod_owner:relabel{%(clusterLabel)s="$cluster", namespace="$namespace", workload="$workload", workload_type=~"$type"}
) by (pod)
Expand Down
10 changes: 10 additions & 0 deletions rules/apps.libsonnet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,16 @@
// Reduces cardinality of this timeseries by #cores, which makes it
// more useable in dashboards. Also, allows us to do things like
// quantile_over_time(...) which would otherwise not be possible.
record: 'node_namespace_pod_container:container_cpu_usage_seconds_total:sum_rate5m',
expr: |||
sum by (%(clusterLabel)s, namespace, pod, container) (
rate(container_cpu_usage_seconds_total{%(cadvisorSelector)s, image!=""}[5m])
) * on (%(clusterLabel)s, namespace, pod) group_left(node) topk by (%(clusterLabel)s, namespace, pod) (
1, max by(%(clusterLabel)s, namespace, pod, node) (kube_pod_info{node!=""})
)
||| % $._config,
},
{
record: 'node_namespace_pod_container:container_cpu_usage_seconds_total:sum_irate',
expr: |||
sum by (%(clusterLabel)s, namespace, pod, container) (
Expand Down
Loading
Loading