[
https://issues.apache.org/jira/browse/HADOOP-17833?focusedWorklogId=779055&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-779055
]
ASF GitHub Bot logged work on HADOOP-17833:
-------------------------------------------
Author: ASF GitHub Bot
Created on: 07/Jun/22 12:25
Start Date: 07/Jun/22 12:25
Worklog Time Spent: 10m
Work Description: steveloughran commented on code in PR #3289:
URL: https://github.com/apache/hadoop/pull/3289#discussion_r891151772
##########
hadoop-tools/hadoop-aws/src/main/java/org/apache/hadoop/fs/s3a/S3AFileSystem.java:
##########
@@ -3970,63 +4062,64 @@ InitiateMultipartUploadResult initiateMultipartUpload(
/**
* Perform post-write actions.
- * <p></p>
+ * <p>
* This operation MUST be called after any PUT/multipart PUT completes
* successfully.
- * <p></p>
- * The actions include:
- * <ol>
- * <li>
- * Calling
- * {@link #deleteUnnecessaryFakeDirectories(Path)}
- * if directory markers are not being retained.
- * </li>
- * <li>
- * Updating any metadata store with details on the newly created
- * object.
- * </li>
- * </ol>
+ * <p>
+ * The actions include calling
+ * {@link #deleteUnnecessaryFakeDirectories(Path)}
+ * if directory markers are not being retained.
* @param key key written to
* @param length total length of file written
* @param eTag eTag of the written object
* @param versionId S3 object versionId of the written object
+ * @param putOptions put object options
*/
@InterfaceAudience.Private
- @Retries.RetryTranslated("Except if failOnMetadataWriteError=false, in which"
- + " case RetryExceptionsSwallowed")
- void finishedWrite(String key, long length, String eTag, String versionId) {
+ @Retries.RetryExceptionsSwallowed
+ void finishedWrite(
+ String key,
+ long length,
+ String eTag,
+ String versionId,
+ PutObjectOptions putOptions) {
LOG.debug("Finished write to {}, len {}. etag {}, version {}",
key, length, eTag, versionId);
- Path p = keyToQualifiedPath(key);
Preconditions.checkArgument(length >= 0, "content length is negative");
- // kick off an async delete
- CompletableFuture<?> deletion;
- if (!keepDirectoryMarkers(p)) {
- deletion = submit(
- unboundedThreadPool, getActiveAuditSpan(),
- () -> {
- deleteUnnecessaryFakeDirectories(
- p.getParent()
- );
- return null;
- });
- } else {
- deletion = null;
+ if (!putOptions.isKeepMarkers()) {
Review Comment:
async was to mix the delete call with the s3guard update for minor
performance benefits. without s3guard, no possibilities of overlapping IO, just
risk of waiting for a spare thread. it has become counterproductive
Issue Time Tracking
-------------------
Worklog Id: (was: 779055)
Time Spent: 9h 40m (was: 9.5h)
> Improve Magic Committer Performance
> -----------------------------------
>
> Key: HADOOP-17833
> URL: https://issues.apache.org/jira/browse/HADOOP-17833
> Project: Hadoop Common
> Issue Type: Improvement
> Components: fs/s3
> Affects Versions: 3.3.1
> Reporter: Steve Loughran
> Assignee: Steve Loughran
> Priority: Minor
> Labels: pull-request-available
> Time Spent: 9h 40m
> Remaining Estimate: 0h
>
> Magic committer tasks can be slow because every file created with
> overwrite=false triggers a HEAD (verify there's no file) and a LIST (that
> there's no dir). And because of delayed manifestations, it may not behave as
> expected.
> ParquetOutputFormat is one example of a library which does this.
> we could fix parquet to use overwrite=true, but (a) there may be surprises in
> other uses (b) it'd still leave the list and (c) do nothing for other formats
> call
> Proposed: createFile() under a magic path to skip all probes for file/dir at
> end of path
> Only a single task attempt Will be writing to that directory and it should
> know what it is doing. If there is conflicting file names and parts across
> tasks that won't even get picked up at this point. Oh and none of the
> committers ever check for this: you'll get the last file manifested (s3a) or
> renamed (file)
> If we skip the checks we will save 2 HTTP requests/file.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
