bstein/titan-iac
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

monitoring: add gpu node fallback

Browse Source
This commit is contained in:
Brad Stein 2025-11-18 10:47:24 -03:00
parent ec76563a86
commit d7e4bcd533
3 changed files with 21 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
scripts/render_dashboards.py
View File

@ -222,18 +222,24 @@ NAMESPACE_CPU_RAW = (
NAMESPACE_RAM_RAW = (
'sum(container_memory_working_set_bytes{namespace!="",pod!="",container!=""}) by (namespace)'
)
NAMESPACE_GPU_RAW = (
GPU_NODES = ["titan-20", "titan-21", "titan-22", "titan-24"]
GPU_NODE_REGEX = "|".join(GPU_NODES)
NAMESPACE_GPU_REQUEST = (
'sum((kube_pod_container_resource_requests{namespace!="",resource="nvidia.com/gpu"}'
' or kube_pod_container_resource_limits{namespace!="",resource="nvidia.com/gpu"})'
') by (namespace)'
' or kube_pod_container_resource_limits{namespace!="",resource="nvidia.com/gpu"})) by (namespace)'
)
NAMESPACE_GPU_WEIGHT = (
"(( "
+ NAMESPACE_GPU_RAW
+ " ) or on(namespace) ( "
+ NAMESPACE_CPU_RAW
+ " * 0))"
NAMESPACE_GPU_FALLBACK = (
'sum by (namespace) (kube_pod_info{namespace!=""}'
f' and on(node) kube_node_info{{node=~"{GPU_NODE_REGEX}"}})'
)
NAMESPACE_GPU_RAW = (
"("
+ NAMESPACE_GPU_REQUEST
+ ") or on(namespace) group_left() ("
+ NAMESPACE_GPU_FALLBACK
+ ")"
)
NAMESPACE_GPU_WEIGHT = NAMESPACE_GPU_RAW
NAMESPACE_COMBINED_FILTER = (
'topk(10, ('
+ NAMESPACE_CPU_RAW

6
services/monitoring/dashboards/atlas-overview.json
View File

@ -722,7 +722,7 @@
},
"targets": [
{
"expr": "100 * ( ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ) or on(namespace) ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) * 0)) * 10)) ) ) / clamp_min(sum( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) ), 1)",
"expr": "100 * ( ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) * 10)) ) ) / clamp_min(sum( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) ), 1)",
"refId": "A",
"legendFormat": "{{namespace}}"
}
@ -764,7 +764,7 @@
},
"targets": [
{
"expr": "100 * ( ( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ) or on(namespace) ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) * 0)) * 10)) ) ) / clamp_min(sum( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ), 1)",
"expr": "100 * ( ( (sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) * 10)) ) ) / clamp_min(sum( (sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) ), 1)",
"refId": "A",
"legendFormat": "{{namespace}}"
}
@ -806,7 +806,7 @@
},
"targets": [
{
"expr": "100 * ( ( sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ) or on(namespace) ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) * 0)) * 10)) ) ) / clamp_min(sum( sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) ), 1)",
"expr": "100 * ( ( sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) * 10)) ) ) / clamp_min(sum( sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) ), 1)",
"refId": "A",
"legendFormat": "{{namespace}}"
}

6
services/monitoring/grafana-dashboard-overview.yaml
View File

@ -731,7 +731,7 @@ data:
},
"targets": [
{
"expr": "100 * ( ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ) or on(namespace) ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) * 0)) * 10)) ) ) / clamp_min(sum( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) ), 1)",
"expr": "100 * ( ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) * 10)) ) ) / clamp_min(sum( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) ), 1)",
"refId": "A",
"legendFormat": "{{namespace}}"
}
@ -773,7 +773,7 @@ data:
},
"targets": [
{
"expr": "100 * ( ( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ) or on(namespace) ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) * 0)) * 10)) ) ) / clamp_min(sum( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ), 1)",
"expr": "100 * ( ( (sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) * 10)) ) ) / clamp_min(sum( (sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) ), 1)",
"refId": "A",
"legendFormat": "{{namespace}}"
}
@ -815,7 +815,7 @@ data:
},
"targets": [
{
"expr": "100 * ( ( sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((( sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace) ) or on(namespace) ( sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace) * 0)) * 10)) ) ) / clamp_min(sum( sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) ), 1)",
"expr": "100 * ( ( sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) ) and on(namespace) ( topk(10, (sum(rate(container_cpu_usage_seconds_total{namespace!=\"\",pod!=\"\",container!=\"\"}[5m])) by (namespace)) + (sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) / 1e9) + ((sum((kube_pod_container_resource_requests{namespace!=\"\",resource=\"nvidia.com/gpu\"} or kube_pod_container_resource_limits{namespace!=\"\",resource=\"nvidia.com/gpu\"})) by (namespace)) or on(namespace) group_left() (sum by (namespace) (kube_pod_info{namespace!=\"\"} and on(node) kube_node_info{node=~\"titan-20|titan-21|titan-22|titan-24\"})) * 10)) ) ) / clamp_min(sum( sum(container_memory_working_set_bytes{namespace!=\"\",pod!=\"\",container!=\"\"}) by (namespace) ), 1)",
"refId": "A",
"legendFormat": "{{namespace}}"
}