Skip to content
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

fix: Optimize Prometheus scrape target discovery #1663

Merged
merged 7 commits into from
Dec 10, 2024
Merged

fix: Optimize Prometheus scrape target discovery #1663

Show file tree
Hide file tree
Changes from 5 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
8b5ccc4
fix prometheus discovery job, optimize discovery for resources on sam…
hisarbalik Dec 7, 2024
60d3751
fix pod affinity for metric agent load test setup
hisarbalik Dec 7, 2024
952e40c
Merge branch 'main' into fix-metric-agent-promethues-scrape-config
hisarbalik Dec 7, 2024
c9c4199
Merge branch 'main' into fix-metric-agent-promethues-scrape-config
hisarbalik Dec 8, 2024
fe62991
Merge branch 'main' into fix-metric-agent-promethues-scrape-config
hisarbalik Dec 8, 2024
d13c836
Merge branch 'main' into fix-metric-agent-promethues-scrape-config
hisarbalik Dec 10, 2024
3da7eaf
move discovery config creation to dedicated function
hisarbalik Dec 10, 2024
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
2 changes: 1 addition & 1 deletion hack/load-tests/metric-agent-test-setup.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50,7 +50,7 @@ spec:
- podAffinityTerm:
labelSelector:
matchExpressions:
- key: app
- key: app.kubernetes.io/name
operator: In
values:
- metric-agent-load-generator
Expand Down
8 changes: 7 additions & 1 deletion internal/otelcollector/config/metric/agent/config.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -187,11 +187,17 @@ type StaticDiscoveryConfig struct {
}

type KubernetesDiscoveryConfig struct {
Role Role `yaml:"role"`
Role Role `yaml:"role"`
Selectors []K8SDiscoverySelector `yaml:"selectors,omitempty"`
}

type Role string

type K8SDiscoverySelector struct {
Role Role `yaml:"role"`
Field string `yaml:"field"`
}

const (
RoleEndpoints Role = "endpoints"
RolePod Role = "pod"
Expand Down
61 changes: 46 additions & 15 deletions internal/otelcollector/config/metric/agent/prometheus_receiver.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,8 +18,9 @@ const (
type AnnotatedResource string

const (
AnnotatedPod AnnotatedResource = "pod"
AnnotatedService AnnotatedResource = "service"
AnnotatedPod AnnotatedResource = "pod"
AnnotatedService AnnotatedResource = "service"
PodNodeSelectorFieldExpression string = "spec.nodeName=${MY_NODE_NAME}"
)

const (
Expand All @@ -35,11 +36,21 @@ func makePrometheusConfigForPods() *PrometheusReceiver {
var config PrometheusReceiver

scrapeConfig := ScrapeConfig{
ScrapeInterval: scrapeInterval,
SampleLimit: sampleLimit,
KubernetesDiscoveryConfigs: []KubernetesDiscoveryConfig{{Role: RolePod}},
JobName: appPodsJobName,
RelabelConfigs: makePrometheusPodsRelabelConfigs(),
ScrapeInterval: scrapeInterval,
SampleLimit: sampleLimit,
KubernetesDiscoveryConfigs: []KubernetesDiscoveryConfig{
{
Role: RolePod,
Selectors: []K8SDiscoverySelector{
{
Role: RolePod,
Field: PodNodeSelectorFieldExpression,
},
},
},
},
k15r marked this conversation as resolved.
Show resolved Hide resolved
JobName: appPodsJobName,
RelabelConfigs: makePrometheusPodsRelabelConfigs(),
}

config.Config.ScrapeConfigs = append(config.Config.ScrapeConfigs, scrapeConfig)
Expand All @@ -55,9 +66,19 @@ func makePrometheusConfigForServices(opts BuildOptions) *PrometheusReceiver {
var config PrometheusReceiver

baseScrapeConfig := ScrapeConfig{
ScrapeInterval: scrapeInterval,
SampleLimit: sampleLimit,
KubernetesDiscoveryConfigs: []KubernetesDiscoveryConfig{{Role: RoleEndpoints}},
ScrapeInterval: scrapeInterval,
SampleLimit: sampleLimit,
KubernetesDiscoveryConfigs: []KubernetesDiscoveryConfig{
{
Role: RoleEndpoints,
Selectors: []K8SDiscoverySelector{
{
Role: RolePod,
Field: PodNodeSelectorFieldExpression,
},
},
},
},
}

httpScrapeConfig := baseScrapeConfig
Expand Down Expand Up @@ -142,11 +163,21 @@ func makePrometheusIstioConfig() *PrometheusReceiver {
Config: PrometheusConfig{
ScrapeConfigs: []ScrapeConfig{
{
JobName: "istio-proxy",
SampleLimit: sampleLimit,
MetricsPath: "/stats/prometheus",
ScrapeInterval: scrapeInterval,
KubernetesDiscoveryConfigs: []KubernetesDiscoveryConfig{{Role: RolePod}},
JobName: "istio-proxy",
SampleLimit: sampleLimit,
MetricsPath: "/stats/prometheus",
ScrapeInterval: scrapeInterval,
KubernetesDiscoveryConfigs: []KubernetesDiscoveryConfig{
{
Role: RolePod,
Selectors: []K8SDiscoverySelector{
{
Role: RolePod,
Field: PodNodeSelectorFieldExpression,
},
},
},
},
RelabelConfigs: []RelabelConfig{
keepIfRunningOnSameNode(NodeAffiliatedPod),
keepIfIstioProxy(),
Expand Down
12 changes: 12 additions & 0 deletions internal/otelcollector/config/metric/agent/testdata/config_istio_enabled.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -175,6 +175,9 @@ receivers:
action: replace
kubernetes_sd_configs:
- role: pod
selectors:
- role: pod
field: spec.nodeName=${MY_NODE_NAME}
prometheus/app-services:
config:
scrape_configs:
Expand Down Expand Up @@ -223,6 +226,9 @@ receivers:
action: replace
kubernetes_sd_configs:
- role: endpoints
selectors:
- role: pod
field: spec.nodeName=${MY_NODE_NAME}
- job_name: app-services-secure
sample_limit: 50000
scrape_interval: 30s
Expand Down Expand Up @@ -268,6 +274,9 @@ receivers:
action: replace
kubernetes_sd_configs:
- role: endpoints
selectors:
- role: pod
field: spec.nodeName=${MY_NODE_NAME}
tls_config:
ca_file: /etc/istio-output-certs/root-cert.pem
cert_file: /etc/istio-output-certs/cert-chain.pem
Expand Down Expand Up @@ -299,6 +308,9 @@ receivers:
action: keep
kubernetes_sd_configs:
- role: pod
selectors:
- role: pod
field: spec.nodeName=${MY_NODE_NAME}
processors:
batch:
send_batch_size: 1024
Expand Down
6 changes: 6 additions & 0 deletions internal/otelcollector/config/metric/agent/testdata/config_istio_not_enabled.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -164,6 +164,9 @@ receivers:
action: replace
kubernetes_sd_configs:
- role: pod
selectors:
- role: pod
field: spec.nodeName=${MY_NODE_NAME}
prometheus/app-services:
config:
scrape_configs:
Expand Down Expand Up @@ -212,6 +215,9 @@ receivers:
action: replace
kubernetes_sd_configs:
- role: endpoints
selectors:
- role: pod
field: spec.nodeName=${MY_NODE_NAME}
processors:
batch:
send_batch_size: 1024
Expand Down
Loading