Details
-
Bug
-
Status: Resolved
-
Major
-
Resolution: Duplicate
-
1.1.0
-
None
-
None
Description
If we enable jvm metrics for executor, master, worker, driver instances, we don't have context where they are coming from ?
This can be a issue if we are collecting all the metrics from different instances are storing into common datastore.
This is mainly running Spark on Yarn but i believe Spark standalone has also this problems.
It would be good if we attach some context for jvm metrics.
Attachments
Issue Links
- duplicates
-
SPARK-5847 Allow for configuring MetricsSystem's use of app ID to namespace all metrics
- Resolved