Skip to content

[SPARK-32916][SHUFFLE][test-maven][test-hadoop2.7] Ensure the number of chunks in meta file and index file are equal #30433

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.

Sign up for GitHub

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

Closed
otterc wants to merge 6 commits into from
Closed

[SPARK-32916][SHUFFLE][test-maven][test-hadoop2.7] Ensure the number of chunks in meta file and index file are equal #30433

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
72c115c
Ensure the number of chunks in meta file and index file are equal
otterc Nov 19, 2020
f5783a9
converting long to byte-array and using RAFile.write(byte[])
otterc Nov 21, 2020
72b5976
Addressed review comments and added another UT
otterc Dec 3, 2020
5422f10
Fixed log statement and comment
otterc Dec 8, 2020
0bce631
Fail pending blocks before writing when number of IOExceptions have e…
otterc Dec 9, 2020
f48d52f
Reuse abortIfNecessary while creating the callback
otterc Dec 9, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 10 additions & 0 deletions common/network-common/src/main/java/org/apache/spark/network/util/TransportConf.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -398,4 +398,14 @@ public long mergedIndexCacheSize() {
return JavaUtils.byteStringAsBytes(
conf.get("spark.shuffle.server.mergedIndexCacheSize", "100m"));
}

/**
* The threshold for number of IOExceptions while merging shuffle blocks to a shuffle partition.
* When the number of IOExceptions while writing to merged shuffle data/index/meta file exceed
* this threshold then the shuffle server will respond back to client to stop pushing shuffle
* blocks for this shuffle partition.
*/
public int ioExceptionsThresholdDuringMerge() {
return conf.getInt("spark.shuffle.server.ioExceptionsThresholdDuringMerge", 4);
}
}
9 changes: 9 additions & 0 deletions common/network-shuffle/src/main/java/org/apache/spark/network/shuffle/ErrorHandler.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -71,6 +71,15 @@ class BlockPushErrorHandler implements ErrorHandler {
public static final String BLOCK_APPEND_COLLISION_DETECTED_MSG_PREFIX =
"Couldn't find an opportunity to write block";

/**
* String constant used for generating exception messages indicating the server encountered
* IOExceptions multiple times, greater than the configured threshold, while trying to merged
* shuffle blocks of the same shuffle partition. When the client receives this this response,
* it will stop pushing any more blocks for the same shuffle partition.
*/
public static final String IOEXCEPTIONS_EXCEEDED_THRESHOLD_PREFIX =
"IOExceptions exceeded the threshold";

@Override
public boolean shouldRetryError(Throwable t) {
// If it is a connection time out or a connection closed exception, no need to retry.
Expand Down
Loading