refine highload rule

This commit is contained in:
Jan Eitzinger 2025-06-27 12:16:17 +02:00
parent 8378784231
commit 983aa592d8

View File

@ -14,13 +14,13 @@
"variables": [ "variables": [
{ {
"name": "load_threshold", "name": "load_threshold",
"expr": "(job.numCores / job.numNodes) * excessivecpuload_threshold_factor" "expr": "cpu_load.limits.peak * excessivecpuload_threshold_factor"
}, },
{ {
"name": "load_perc", "name": "load_perc",
"expr": "1.0 - (cpu_load.avg / cpu_load.limits.peak)" "expr": "1.0 - (cpu_load.avg / cpu_load.limits.peak)"
} }
], ],
"rule": "cpu_load.avg > cpu_load.limits.peak", "rule": "cpu_load.avg > load_threshold",
"hint": "This job was detected as excessiveload because the average cpu load {{.cpu_load.avg}} falls above the threshold {{.cpu_load.limits.peak}}." "hint": "This job was detected as excessiveload because the average cpu load {{.cpu_load.avg}} falls above the threshold {{.load_threshold}}."
} }