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This repository has been archived by the owner on Apr 27, 2022. It is now read-only.
This is an issue to brainstorm about possible uses of spark.metrics.namespace, as described here
A possible example. Set spark.metrics.namespace to the spark cluster name assigned by the user or oshinko. So for a cluster named "wilma" the metrics from driver and executors would be:
wilma.metric.name instead of spark_app_uid.metric.name
The text was updated successfully, but these errors were encountered:
i feel like this is something that app developers should be doing in their code. it might be nice to have a default back to the cluster name, but i'm not sure how we would avoid stomping on the value provided by an app in that case.
This is an issue to brainstorm about possible uses of
spark.metrics.namespace
, as described hereA possible example. Set
spark.metrics.namespace
to the spark cluster name assigned by the user or oshinko. So for a cluster named "wilma" the metrics from driver and executors would be:wilma.metric.name
instead ofspark_app_uid.metric.name
The text was updated successfully, but these errors were encountered: