TEZ-4547: Add Tez AM JobID to the JobConf #339
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
Some committers require a job-wide UUID to function correctly. Adding the AM JobID to the JobConf will allow applications to pass that to the committers that need it.
VenkatSNarayanan
force-pushed
the
vnarayanan/TEZ-4547
branch
from
25cae6d
to
9040a2f
Compare
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
r0hini
reviewed
Mar 13, 2024
| @@ -417,6 +418,7 @@ protected List<Event> initializeBase() throws IOException, InterruptedException | |||
| .createMockTaskAttemptID(getContext().getApplicationId().getClusterTimestamp(), | |||
| getContext().getTaskVertexIndex(), getContext().getApplicationId().getId(), | |||
| getContext().getTaskIndex(), getContext().getTaskAttemptNumber(), isMapperOutput); | |||
| jobConf.set(MRJobConfig.MR_PARENT_JOB_ID, new JobID(String.valueOf(getContext().getApplicationId().getClusterTimestamp()), getContext().getApplicationId().getId()).toString()); | |||
There was a problem hiding this comment.
Move this to line above TaskAttemptID taskAttemptId =
r0hini
reviewed
Mar 13, 2024
Comment on lines
151
to
152
| assertNotEquals(parentJobID,invalidJobID); | ||
| assertNotEquals(output.jobConf.get(org.apache.hadoop.mapred.JobContext.TASK_ATTEMPT_ID),parentJobID); |
There was a problem hiding this comment.
Fix code formatting. space after ,
shameersss1
suggested changes
Mar 17, 2024
| @@ -119,6 +119,7 @@ public void abortOutput(VertexStatus.State finalState) throws IOException { | |||
| || jobConf.getBoolean("mapred.mapper.new-api", false)) { | |||
| newApiCommitter = true; | |||
| } | |||
| jobConf.set(MRJobConfig.MR_PARENT_JOB_ID, new org.apache.hadoop.mapred.JobID(String.valueOf(getContext().getApplicationId().getClusterTimestamp()), getContext().getApplicationId().getId()).toString()); | |||
There was a problem hiding this comment.
Can we have String.valueOf(getContext().getApplicationId().getClusterTimestamp()), getContext().getApplicationId().getId()).toString( inside a method and re-use it in MROutput.java as well ?
Yes, that was the plan. The property name was just chosen arbitrarily so I could put the PR up, any suggestions for a better one are welcome. |
This commit also adds the DAG identifier to the job UUID to ensure that multiple jobs within the same session will be assigned different UUIDs.
This comment was marked as outdated.
This comment was marked as outdated.
shameersss1
suggested changes
Mar 20, 2024
tez-mapreduce/src/main/java/org/apache/tez/mapreduce/committer/MROutputCommitter.java
Outdated
Show resolved
Hide resolved
tez-mapreduce/src/main/java/org/apache/tez/mapreduce/common/Utils.java
Outdated
Show resolved
Hide resolved
tez-mapreduce/src/main/java/org/apache/tez/mapreduce/hadoop/MRJobConfig.java
Outdated
Show resolved
Hide resolved
tez-mapreduce/src/main/java/org/apache/tez/mapreduce/output/MROutput.java
Outdated
Show resolved
Hide resolved
This comment was marked as outdated.
This comment was marked as outdated.
|
@shameersss1 We could actually just set fs.s3a.committer.uuid directly instead of the indirection through the other setting. |
Switch UUID property name to the one required by S3A committers.
This comment was marked as outdated.
This comment was marked as outdated.
|
@abstractdog - Could you please review the same ? |
Refactors the implementation to reuse Tez's DAGID type instead of hand-rolling our own.
|
🎊 +1 overall
This message was automatically generated. |
|
@abstractdog @shameersss1 Is there anything else needed? |
abstractdog
reviewed
May 14, 2024
tez-mapreduce/src/main/java/org/apache/tez/mapreduce/hadoop/MRJobConfig.java
Outdated
Show resolved
Hide resolved
abstractdog
reviewed
May 14, 2024
tez-dag/src/main/java/org/apache/tez/dag/app/dag/impl/OutputCommitterContextImpl.java
Outdated
Show resolved
Hide resolved
|
🎊 +1 overall
This message was automatically generated. |
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.
Some committers require a job-wide UUID to function correctly. Adding the AM JobID to the JobConf
will allow applications to pass that to
the committers that need it.