[SPARK-8112][Streaming]Fix the negative event count issue #6659
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
|
Test build #34223 has finished for PR 6659 at commit
|
Contributor
There was a problem hiding this comment.
You could ensure this in a require in the body of the case class.
Contributor
|
LGTM. We can test this in PR #6614 |
|
Test build #34250 has finished for PR 6659 at commit
|
asfgit
pushed a commit
that referenced
this pull request
Jun 5, 2015
Author: zsxwing <[email protected]> Closes #6659 from zsxwing/SPARK-8112 and squashes the following commits: a5d7da6 [zsxwing] Address comments d255b6e [zsxwing] Fix the negative event count issue (cherry picked from commit 4f16d3f) Signed-off-by: Tathagata Das <[email protected]>
jeanlyn
pushed a commit
to jeanlyn/spark
that referenced
this pull request
Jun 12, 2015
Author: zsxwing <[email protected]> Closes apache#6659 from zsxwing/SPARK-8112 and squashes the following commits: a5d7da6 [zsxwing] Address comments d255b6e [zsxwing] Fix the negative event count issue
asfgit
pushed a commit
that referenced
this pull request
Jun 19, 2015
…orrect count at Spark UI tdas zsxwing this is the new PR for Spark-8080 I have merged #6659 Also to mention , for MEMORY_ONLY settings , when Block is not able to unrollSafely to memory if enough space is not there, BlockManager won't try to put the block and ReceivedBlockHandler will throw SparkException as it could not find the block id in PutResult. Thus number of records in block won't be counted if Block failed to unroll in memory. Which is fine. For MEMORY_DISK settings , if BlockManager not able to unroll block to memory, block will still get deseralized to Disk. Same for WAL based store. So for those cases ( storage level = memory + disk ) number of records will be counted even though the block not able to unroll to memory. thus I added the isFullyConsumed in the CountingIterator but have not used it as such case will never happen that block not fully consumed and ReceivedBlockHandler still get the block ID. I have added few test cases to cover those block unrolling scenarios also. Author: Dibyendu Bhattacharya <[email protected]> Author: U-PEROOT\UBHATD1 <[email protected]> Closes #6707 from dibbhatt/master and squashes the following commits: f6cb6b5 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI f37cfd8 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 5a8344a [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI Count ByteBufferBlock as 1 count fceac72 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 0153e7e [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI Fixed comments given by @zsxwing 4c5931d [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 01e6dc8 [U-PEROOT\UBHATD1] A
asfgit
pushed a commit
that referenced
this pull request
Jun 19, 2015
…orrect count at Spark UI tdas zsxwing this is the new PR for Spark-8080 I have merged #6659 Also to mention , for MEMORY_ONLY settings , when Block is not able to unrollSafely to memory if enough space is not there, BlockManager won't try to put the block and ReceivedBlockHandler will throw SparkException as it could not find the block id in PutResult. Thus number of records in block won't be counted if Block failed to unroll in memory. Which is fine. For MEMORY_DISK settings , if BlockManager not able to unroll block to memory, block will still get deseralized to Disk. Same for WAL based store. So for those cases ( storage level = memory + disk ) number of records will be counted even though the block not able to unroll to memory. thus I added the isFullyConsumed in the CountingIterator but have not used it as such case will never happen that block not fully consumed and ReceivedBlockHandler still get the block ID. I have added few test cases to cover those block unrolling scenarios also. Author: Dibyendu Bhattacharya <[email protected]> Author: U-PEROOT\UBHATD1 <[email protected]> Closes #6707 from dibbhatt/master and squashes the following commits: f6cb6b5 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI f37cfd8 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 5a8344a [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI Count ByteBufferBlock as 1 count fceac72 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 0153e7e [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI Fixed comments given by @zsxwing 4c5931d [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 01e6dc8 [U-PEROOT\UBHATD1] A
nemccarthy
pushed a commit
to nemccarthy/spark
that referenced
this pull request
Jun 19, 2015
Author: zsxwing <[email protected]> Closes apache#6659 from zsxwing/SPARK-8112 and squashes the following commits: a5d7da6 [zsxwing] Address comments d255b6e [zsxwing] Fix the negative event count issue
nemccarthy
pushed a commit
to nemccarthy/spark
that referenced
this pull request
Jun 19, 2015
…orrect count at Spark UI tdas zsxwing this is the new PR for Spark-8080 I have merged apache#6659 Also to mention , for MEMORY_ONLY settings , when Block is not able to unrollSafely to memory if enough space is not there, BlockManager won't try to put the block and ReceivedBlockHandler will throw SparkException as it could not find the block id in PutResult. Thus number of records in block won't be counted if Block failed to unroll in memory. Which is fine. For MEMORY_DISK settings , if BlockManager not able to unroll block to memory, block will still get deseralized to Disk. Same for WAL based store. So for those cases ( storage level = memory + disk ) number of records will be counted even though the block not able to unroll to memory. thus I added the isFullyConsumed in the CountingIterator but have not used it as such case will never happen that block not fully consumed and ReceivedBlockHandler still get the block ID. I have added few test cases to cover those block unrolling scenarios also. Author: Dibyendu Bhattacharya <[email protected]> Author: U-PEROOT\UBHATD1 <[email protected]> Closes apache#6707 from dibbhatt/master and squashes the following commits: f6cb6b5 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI f37cfd8 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 5a8344a [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI Count ByteBufferBlock as 1 count fceac72 [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 0153e7e [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI Fixed comments given by @zsxwing 4c5931d [Dibyendu Bhattacharya] [SPARK-8080][STREAMING] Receiver.store with Iterator does not give correct count at Spark UI 01e6dc8 [U-PEROOT\UBHATD1] A
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
No description provided.