LucaCanali commented on a change in pull request #23838: [SPARK-26928][CORE]
Add driver CPU Time to the metrics system
URL: https://github.com/apache/spark/pull/23838#discussion_r260961195
##########
File path: core/src/main/scala/org/apache/spark/SparkContext.scala
##########
@@ -568,6 +569,10 @@ class SparkContext(config: SparkConf) extends Logging {
_taskScheduler.postStartHook()
_env.metricsSystem.registerSource(_dagScheduler.metricsSource)
_env.metricsSystem.registerSource(new
BlockManagerSource(_env.blockManager))
+ // Register JVMCPUSource to the metrics system if
spark.metrics.cpu.time.driver.enabled==true
Review comment:
Good point. Actually, I would not mind to have this metrics gauge always on
for the driver too. I don't think the overhead is high. However, I was just
being cautious here following previous discussions on the implementation of
this gauge for the executor instance [SPARK-25228]. I would like to ping
@srowen for comment as he helped back then with [SPARK-25228].
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
