You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Jun 6, 2024. It is now read-only.
Current result for query of yarn_gpus_used is yarn_gpus_used{cluster="cluster_0",instance="xx.xx.xx.xx:9459",job="pai_serivce_exporter",pai_service_name="hadoop-resource-manager",scraped_from="hadoop-resource-manager-ds-9tkc4"}
What would you like to be added:
emit
yarn_gpus_used
with each node's info.Why is this needed:
Current result for query of
yarn_gpus_used
isyarn_gpus_used{cluster="cluster_0",instance="xx.xx.xx.xx:9459",job="pai_serivce_exporter",pai_service_name="hadoop-resource-manager",scraped_from="hadoop-resource-manager-ds-9tkc4"}
this metrics from http://xx.xx.xx.xx:8088/ws/v1/cluster/metrics , should change it to http://xx.xx.xx.xx:8088/ws/v1/cluster/nodes, which has more information, should be good for debugging and calculating free gpu of each node.
Without this feature, how does the current module work:
Can not break down
yarn_gpus_used
Components that may involve changes:
yarn-exporter
The text was updated successfully, but these errors were encountered: