-
Notifications
You must be signed in to change notification settings - Fork 29k
[SPARK-23472][CORE] Add defaultJavaOptions for driver and executor. #24804
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
|
cc @vanzin @squito @rdblue since you were involved in the previous PR. I've considered many options but this looks the best from my perspective. |
|
Test build #106196 has finished for PR 24804 at commit
|
|
You'll need to add similar logic to |
|
Somehow I had the feeling this code part works with SparkConf and default fetched with extra (which is wrong). Not sure how this can be made more generic. |
|
Test build #106250 has finished for PR 24804 at commit
|
|
Test build #106251 has finished for PR 24804 at commit
|
vanzin
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Looks ok.
core/src/main/scala/org/apache/spark/internal/config/package.scala
Outdated
Show resolved
Hide resolved
core/src/main/scala/org/apache/spark/internal/config/package.scala
Outdated
Show resolved
Hide resolved
core/src/test/scala/org/apache/spark/internal/config/ConfigEntrySuite.scala
Show resolved
Hide resolved
|
BTW it would be good to double check the places where cluster-mode drivers are started and make sure they also pick up the new config. e.g. Which I think would completely ignore the stuff you're adding here... |
* Removed unused config entries * Fallback test added * Fixed Mesos and Rest server property handling
|
Test build #106934 has finished for PR 24804 at commit
|
|
Test build #106941 has finished for PR 24804 at commit
|
squito
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
looks ok, though I haven't looked at whether this is getting used everywhere it needs to be yet (marcelo's earlier comment)
core/src/main/scala/org/apache/spark/internal/config/ConfigEntry.scala
Outdated
Show resolved
Hide resolved
|
Test build #106993 has finished for PR 24804 at commit
|
|
retest this please |
|
Test build #106998 has finished for PR 24804 at commit
|
|
Test build #107017 has finished for PR 24804 at commit
|
vanzin
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Small doc nit; there's also MesosCoarseGrainedSchedulerBackend.scala which seems easy to fix to add support for this. There might be another spot in standalone that may need a change, but I'm having a hard time following that code... if someone cares about it they can figure it out, I guess.
I've double checked the mentioned file but as I see it uses and |
|
Test build #107067 has finished for PR 24804 at commit
|
You're right, I must have misread it. |
|
retest this please |
|
Test build #107489 has finished for PR 24804 at commit
|
|
Merging to master. |
|
This is great! Thank you for fixing on this @gaborgsomogyi and for reviewing @vanzin and @squito! |
## What changes were proposed in this pull request? This PR adds two new config properties: `spark.driver.defaultJavaOptions` and `spark.executor.defaultJavaOptions`. These are intended to be set by administrators in a file of defaults for options like JVM garbage collection algorithm. Users will still set `extraJavaOptions` properties, and both sets of JVM options will be added to start a JVM (default options are prepended to extra options). ## How was this patch tested? Existing + additional unit tests. ``` cd docs/ SKIP_API=1 jekyll build ``` Manual webpage check. Closes apache#24804 from gaborgsomogyi/SPARK-23472. Authored-by: Gabor Somogyi <[email protected]> Signed-off-by: Marcelo Vanzin <[email protected]>
This PR adds two new config properties: `spark.driver.defaultJavaOptions` and `spark.executor.defaultJavaOptions`. These are intended to be set by administrators in a file of defaults for options like JVM garbage collection algorithm. Users will still set `extraJavaOptions` properties, and both sets of JVM options will be added to start a JVM (default options are prepended to extra options). Existing + additional unit tests. ``` cd docs/ SKIP_API=1 jekyll build ``` Manual webpage check. Closes apache#24804 from gaborgsomogyi/SPARK-23472. Authored-by: Gabor Somogyi <[email protected]> Signed-off-by: Marcelo Vanzin <[email protected]> Ref: LIHADOOP-55812 RB=2325692 BUG=LIHADOOP-55812 G=spark-reviewers A=ekrogen
What changes were proposed in this pull request?
This PR adds two new config properties:
spark.driver.defaultJavaOptionsandspark.executor.defaultJavaOptions. These are intended to be set by administrators in a file of defaults for options like JVM garbage collection algorithm. Users will still setextraJavaOptionsproperties, and both sets of JVM options will be added to start a JVM (default options are prepended to extra options).How was this patch tested?
Existing + additional unit tests.
Manual webpage check.