Skip to content

Commit 69d0953

Browse files
committed
fix(grafana): replace phantom metrics with standard cAdvisor/kube-metrics queries
Remove nonexistent 'nantian_gateway_dataplane_container_*' metrics from all panel queries. The dataplane does not emit container-level metrics. Replace with standard cAdvisor (container_cpu_usage_seconds_total, container_memory_working_set_bytes, container_cpu_cfs_*) and kube-state-metrics (kube_pod_container_resource_*) equivalents. Also fix: - xDS ACK/NACK: remove phantom 'xds_publish_nack_lag_seconds' metric, use ack_lag histogram for both panels - Add job label filters to xDS and failover queries - Replace hardcoded pod=~'nantian-gw-dataplane-.*' with $pod_dp variable - Add 'or vector(0)' to failover gauge to prevent empty panels
1 parent 1a1247e commit 69d0953

1 file changed

Lines changed: 11 additions & 16 deletions

File tree

deploy/observability/grafana/pingora-gateway-observability-dashboard.json

Lines changed: 11 additions & 16 deletions
Original file line numberDiff line numberDiff line change
@@ -344,7 +344,7 @@
344344
},
345345
"targets": [
346346
{
347-
"expr": "sum(histogram_quantile(0.95, rate(nantian_gateway_controlplane_xds_publish_ack_lag_seconds_bucket[5m])))",
347+
"expr": "histogram_quantile(0.95, sum(rate(nantian_gateway_controlplane_xds_publish_ack_lag_seconds_bucket{job=~\"$job_controlplane\"}[5m])) by (le))",
348348
"refId": "A"
349349
}
350350
],
@@ -379,7 +379,7 @@
379379
},
380380
"targets": [
381381
{
382-
"expr": "sum(nantian_gateway_dataplane_container_cpu_throttle_ratio{namespace=~\"$namespace\"} or rate(container_cpu_cfs_throttled_periods_total{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\"}[5m]) / rate(container_cpu_cfs_periods_total{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\"}[5m]))",
382+
"expr": "sum(rate(container_cpu_cfs_throttled_periods_total{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\"}[5m]) / rate(container_cpu_cfs_periods_total{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\"}[5m]))",
383383
"refId": "A"
384384
}
385385
],
@@ -414,7 +414,7 @@
414414
},
415415
"targets": [
416416
{
417-
"expr": "sum(nantian_gateway_dataplane_container_memory_working_set_bytes{namespace=~\"$namespace\"} or container_memory_working_set_bytes{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\"})",
417+
"expr": "sum(container_memory_working_set_bytes{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\"})",
418418
"refId": "A"
419419
}
420420
],
@@ -833,7 +833,7 @@
833833
},
834834
"targets": [
835835
{
836-
"expr": "sum(nantian_gateway_dataplane_traffic_failover_success_rate)",
836+
"expr": "sum(nantian_gateway_dataplane_traffic_failover_success_rate{job=~\"$job_dataplane\"}) or vector(0)",
837837
"refId": "A"
838838
}
839839
],
@@ -1187,12 +1187,12 @@
11871187
"options": {},
11881188
"targets": [
11891189
{
1190-
"expr": "nantian_gateway_dataplane_container_cpu_cores{namespace=~\"$namespace\"} or rate(container_cpu_usage_seconds_total{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\"}[5m])",
1190+
"expr": "rate(container_cpu_usage_seconds_total{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\"}[5m])",
11911191
"legendFormat": "cores",
11921192
"refId": "A"
11931193
},
11941194
{
1195-
"expr": "nantian_gateway_dataplane_container_cpu_request_cores{namespace=~\"$namespace\"} or avg by (pod) (kube_pod_container_resource_requests{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\",resource=\"cpu\"})",
1195+
"expr": "avg by (pod) (kube_pod_container_resource_requests{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\",resource=\"cpu\"})",
11961196
"legendFormat": "request",
11971197
"refId": "B"
11981198
}
@@ -1220,7 +1220,7 @@
12201220
"options": {},
12211221
"targets": [
12221222
{
1223-
"expr": "nantian_gateway_dataplane_container_cpu_throttle_ratio{namespace=~\"$namespace\"} or rate(container_cpu_cfs_throttled_periods_total{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\"}[5m]) / rate(container_cpu_cfs_periods_total{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\"}[5m])",
1223+
"expr": "rate(container_cpu_cfs_throttled_periods_total{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\"}[5m]) / rate(container_cpu_cfs_periods_total{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\"}[5m])",
12241224
"legendFormat": "throttle",
12251225
"refId": "A"
12261226
}
@@ -1248,17 +1248,17 @@
12481248
"options": {},
12491249
"targets": [
12501250
{
1251-
"expr": "sum by (pod) (nantian_gateway_dataplane_container_memory_working_set_bytes{namespace=~\"$namespace\"} or container_memory_working_set_bytes{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\"})",
1251+
"expr": "sum by (pod) (container_memory_working_set_bytes{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\"})",
12521252
"legendFormat": "{{pod}}",
12531253
"refId": "A"
12541254
},
12551255
{
1256-
"expr": "max by (pod) (nantian_gateway_dataplane_container_memory_limit_bytes{namespace=~\"$namespace\"} or avg by (pod) (kube_pod_container_resource_limits{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\",resource=\"memory\"}))",
1256+
"expr": "avg by (pod) (kube_pod_container_resource_limits{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\",resource=\"memory\"})",
12571257
"legendFormat": "limit",
12581258
"refId": "B"
12591259
},
12601260
{
1261-
"expr": "max by (pod) (nantian_gateway_dataplane_container_memory_request_bytes{namespace=~\"$namespace\"} or avg by (pod) (kube_pod_container_resource_requests{namespace=~\"$namespace\",pod=~\"nantian-gw-dataplane-.*\",resource=\"memory\"}))",
1261+
"expr": "avg by (pod) (kube_pod_container_resource_requests{namespace=~\"$namespace\",container!=\"\",pod=~\"$pod_dp\",resource=\"memory\"})",
12621262
"legendFormat": "request",
12631263
"refId": "C"
12641264
}
@@ -1474,14 +1474,9 @@
14741474
"options": {},
14751475
"targets": [
14761476
{
1477-
"expr": "histogram_quantile(0.99, rate(nantian_gateway_controlplane_xds_publish_ack_lag_seconds_bucket[5m]))",
1477+
"expr": "histogram_quantile(0.99, sum(rate(nantian_gateway_controlplane_xds_publish_ack_lag_seconds_bucket{job=~\"$job_controlplane\"}[5m])) by (le))",
14781478
"legendFormat": "ACK p99",
14791479
"refId": "A"
1480-
},
1481-
{
1482-
"expr": "histogram_quantile(0.99, rate(nantian_gateway_controlplane_xds_publish_nack_lag_seconds_bucket[5m]))",
1483-
"legendFormat": "NACK p99",
1484-
"refId": "B"
14851480
}
14861481
],
14871482
"title": "xDS ACK/NACK Lag",

0 commit comments

Comments
 (0)