[spark] ray on spark creates spark job using stage scheduling #31397
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
WeichenXu123
requested review from
architkulkarni,
wuisawesome,
DmitriGekhtman,
maxpumperla,
pcmoritz and
a team
as code owners
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
WeichenXu123
commented
Jan 9, 2023
Comment on lines
+56
to
+57
| # TODO: Replace this with pyspark==3.4 once it is released. | ||
| https://ml-team-public-read.s3.us-west-2.amazonaws.com/pyspark-3.4.0.dev0.tar.gz |
There was a problem hiding this comment.
Note:
Spark stage scheduling feature (on standalone mode spark cluster) is introduced in apache/spark 3.4 (it will be released this month)
So for testing purpose, I built a package from apache/spark master
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
jjyao
reviewed
Jan 11, 2023
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
amogkam
reviewed
Jan 14, 2023
There was a problem hiding this comment.
Thanks @weichen123! Overall lgtm- left some small comments.
Will leave to @jjyao for final approval and marge.
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
7 tasks
jjyao
reviewed
Jan 16, 2023
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
jjyao
approved these changes
Jan 17, 2023
| ) | ||
|
|
||
|
|
||
| def get_avail_mem_per_ray_worker_node( |
There was a problem hiding this comment.
can be addressed in a follow-up PR: we should add type annotations even to private methods.
|
You will also need @ericl approval for doc changes. |
harupy
reviewed
Jan 18, 2023
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
Signed-off-by: Weichen Xu <weichen.xu@databricks.com>
|
CC @ericl Could you approve the doc part of the PR ? Thank you! |
|
CC @ericl Would you help approve the doc changes in the PR? Thank you! |
maxpumperla
approved these changes
Jan 23, 2023
|
Failed tests are unrelated. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Why are these changes needed?
ray on spark creates spark job using stage scheduling, so that ray cluster spark job can use different task resources config ( spark.task.cpus / spark.task.resource.gpu.amount ), otherwise it has to use spark application level config, which is inconvenient on Databricks. 2 new arguments are added:
num_cpus_per_nodeandnum_gpus_per_nodeimprove ray worker memory allocation computation.
refactor
_init_ray_clusterinterface, make it fit better for instrumentation logging patching (make arguments key value only, and adjust some arguments, make all arguments to be validated values)Related issue number
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.