ODIn/spark-instrumented-optimizer
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

[SPARK-17715][SCHEDULER] Make task launch logs DEBUG

Browse source 
## What changes were proposed in this pull request?

Ramp down the task launch logs from INFO to DEBUG. Task launches can happen orders of magnitude more than executor registration so it makes the logs easier to handle if they are different log levels. For larger jobs, there can be 100,000s of task launches which makes the driver log huge.

## How was this patch tested?

No tests, as this is a trivial change.

Author: Brian Cho <bcho@fb.com>

Closes #15290 from dafrista/ramp-down-task-logging.
This commit is contained in:
Brian Cho 2016-09-29 15:59:17 -04:00 committed by Andrew Or
parent cb87b3ced9
commit 027dea8f29
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
core/src/main/scala/org/apache/spark/scheduler/cluster/CoarseGrainedSchedulerBackend.scala
View file

@ -265,7 +265,7 @@ class CoarseGrainedSchedulerBackend(scheduler: TaskSchedulerImpl, val rpcEnv: Rp
val executorData = executorDataMap(task.executorId)
executorData.freeCores -= scheduler.CPUS_PER_TASK
logInfo(s"Launching task ${task.taskId} on executor id: ${task.executorId} hostname: " +
logDebug(s"Launching task ${task.taskId} on executor id: ${task.executorId} hostname: " +
s"${executorData.executorHost}.")
executorData.executorEndpoint.send(LaunchTask(new SerializableBuffer(serializedTask)))