[SPARK-26140] Enable custom metrics implementation in shuffle reader #23105
Conversation
This file contains hidden or 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
rxin
changed the title
rxin
commented
Nov 21, 2018
| * shuffle dependency, and all temporary metrics will be merged into the [[ShuffleReadMetrics]] at | ||
| * last. | ||
| */ | ||
| private[spark] class TempShuffleReadMetrics { |
Contributor
Author
There was a problem hiding this comment.
this was moved to TempShuffleReadMetrics
rxin
changed the title
Contributor
Author
|
Test build #99128 has finished for PR 23105 at commit
|
|
Test build #99131 has finished for PR 23105 at commit
|
|
Test build #99129 has finished for PR 23105 at commit
|
cloud-fan
reviewed
Nov 23, 2018
| package org.apache.spark.shuffle | ||
|
|
||
| /** | ||
| * An interface for reporting shuffle information, for each shuffle. This interface assumes |
Contributor
There was a problem hiding this comment.
for each shuffle -> for each reducer of a shuffle?
cloud-fan
reviewed
Nov 23, 2018
| * All methods have additional Spark visibility modifier to allow public, concrete implementations | ||
| * that still have these methods marked as private[spark]. | ||
| */ | ||
| private[spark] trait ShuffleReadMetricsReporter { |
Contributor
There was a problem hiding this comment.
how do we plan to use this interface later on? It's not used in this PR.
Contributor
Author
There was a problem hiding this comment.
@xuanyuanking just submitted a PR on how to use it :)
cloud-fan
reviewed
Nov 23, 2018
| endPartition: Int, | ||
| context: TaskContext): ShuffleReader[K, C] | ||
| context: TaskContext, | ||
| metrics: ShuffleMetricsReporter): ShuffleReader[K, C] |
Contributor
There was a problem hiding this comment.
IIUC, we should pass a read metrics reporter here, as this method is getReader which is called by the reducers to read shuffle files.
Contributor
Author
There was a problem hiding this comment.
It is a read metrics here actually. In the write PR this is renamed ShuffleReadMetricsReporter.
gatorsmile
reviewed
Nov 23, 2018
Member
gatorsmile
left a comment
gatorsmile
left a comment
There was a problem hiding this comment.
LGTM
Thanks! Merged to master.
asfgit
pushed a commit
that referenced
this pull request
Nov 27, 2018
## What changes were proposed in this pull request? In #23105, due to working on two parallel PRs at once, I made the mistake of committing the copy of the PR that used the name ShuffleMetricsReporter for the interface, rather than the appropriate one ShuffleReadMetricsReporter. This patch fixes that. ## How was this patch tested? This should be fine as long as compilation passes. Closes #23147 from rxin/ShuffleReadMetricsReporter. Authored-by: Reynold Xin <[email protected]> Signed-off-by: gatorsmile <[email protected]>
jackylee-ch
pushed a commit
to jackylee-ch/spark
that referenced
this pull request
Feb 18, 2019
## What changes were proposed in this pull request? This patch defines an internal Spark interface for reporting shuffle metrics and uses that in shuffle reader. Before this patch, shuffle metrics is tied to a specific implementation (using a thread local temporary data structure and accumulators). After this patch, callers that define their own shuffle RDDs can create a custom metrics implementation. With this patch, we would be able to create a better metrics for the SQL layer, e.g. reporting shuffle metrics in the SQL UI, for each exchange operator. Note that I'm separating read side and write side implementations, as they are very different, to simplify code review. Write side change is at apache#23106 ## How was this patch tested? No behavior change expected, as it is a straightforward refactoring. Updated all existing test cases. Closes apache#23105 from rxin/SPARK-26140. Authored-by: Reynold Xin <[email protected]> Signed-off-by: gatorsmile <[email protected]>
jackylee-ch
pushed a commit
to jackylee-ch/spark
that referenced
this pull request
Feb 18, 2019
## What changes were proposed in this pull request? In apache#23105, due to working on two parallel PRs at once, I made the mistake of committing the copy of the PR that used the name ShuffleMetricsReporter for the interface, rather than the appropriate one ShuffleReadMetricsReporter. This patch fixes that. ## How was this patch tested? This should be fine as long as compilation passes. Closes apache#23147 from rxin/ShuffleReadMetricsReporter. Authored-by: Reynold Xin <[email protected]> Signed-off-by: gatorsmile <[email protected]>
jackylee-ch
pushed a commit
to jackylee-ch/spark
that referenced
this pull request
Feb 18, 2019
## What changes were proposed in this pull request? This is the write side counterpart to apache#23105 ## How was this patch tested? No behavior change expected, as it is a straightforward refactoring. Updated all existing test cases. Closes apache#23106 from rxin/SPARK-26141. Authored-by: Reynold Xin <[email protected]> Signed-off-by: Reynold Xin <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
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.
What changes were proposed in this pull request?
This patch defines an internal Spark interface for reporting shuffle metrics and uses that in shuffle reader. Before this patch, shuffle metrics is tied to a specific implementation (using a thread local temporary data structure and accumulators). After this patch, callers that define their own shuffle RDDs can create a custom metrics implementation.
With this patch, we would be able to create a better metrics for the SQL layer, e.g. reporting shuffle metrics in the SQL UI, for each exchange operator.
Note that I'm separating read side and write side implementations, as they are very different, to simplify code review. Write side change is at #23106
How was this patch tested?
No behavior change expected, as it is a straightforward refactoring. Updated all existing test cases.