HADOOP-18805. S3A prefetch tests to work with small files #5851
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
Contributor
Author
|
This PR build is also failing with likely jenkins env issue https://ci-hadoop.apache.org/job/hadoop-multibranch/job/PR-5851/1/console |
This was referenced Jul 17, 2023
Contributor
Author
|
jenkins build issue resolved as part of INFRA-24797 |
|
🎊 +1 overall
This message was automatically generated. |
Contributor
steveloughran
left a comment
steveloughran
left a comment
There was a problem hiding this comment.
I'd also imagined this jira would cover the original large file test, as it has similar issues. we shouldn't need to work with landsat at all.
| S3ATestUtils.removeBaseAndBucketOverrides(conf, PREFETCH_BLOCK_SIZE_KEY); | ||
| conf.setBoolean(PREFETCH_ENABLED_KEY, true); | ||
| conf.setInt(PREFETCH_MAX_BLOCKS_COUNT, Integer.parseInt(maxBlocks)); | ||
| conf.setInt(PREFETCH_BLOCK_SIZE_KEY, S_1K * 10); |
Contributor
There was a problem hiding this comment.
given this is fixed, why have blocksize reread on L115?
proposed: make blocksize constant BLOCK_SIZE, use as appropriate
Contributor
Author
you mean |
Contributor
Author
|
removed dependency on landsat for prefetch tests, re-tested the whole suite against |
steveloughran
requested changes
Jul 19, 2023
Contributor
steveloughran
left a comment
steveloughran
left a comment
There was a problem hiding this comment.
this is good. now we have small filesystems everywhere we can actually skip creating new filesystems in test cases, just use getFileSystem().
this will make the classes even simpler
| S3ATestUtils.removeBaseAndBucketOverrides(conf, PREFETCH_ENABLED_KEY); | ||
| S3ATestUtils.removeBaseAndBucketOverrides(conf, PREFETCH_BLOCK_SIZE_KEY); | ||
| conf.setBoolean(PREFETCH_ENABLED_KEY, true); | ||
| conf.setInt(PREFETCH_BLOCK_SIZE_KEY, BLOCK_SIZE); |
Contributor
There was a problem hiding this comment.
because the size is now being set in this method, the filesystem automatically created/destroyed is already set up for the tests; we can replace openFS() entirely.
Contributor
Author
There was a problem hiding this comment.
done, makes them much simpler
| } | ||
|
|
||
| private void openFS() throws Exception { | ||
| private void openFS(String fileName) throws Exception { |
Contributor
There was a problem hiding this comment.
change to using the existing fs, and using methodPath() to generate a per test case path (no test uses more than one file, does it?)
void createLargeFile() throws IOException {
Configuration conf = getConfiguration();
byte[] data = ContractTestUtils.dataset(S_1K * 72, 'x', 26);
largeFile = methodPath();
FileSystem largeFileFS = getFileSystem();
ContractTestUtils.writeDataset(largeFileFS, largeFile, data, data.length, 16, true);
FileStatus fileStatus = largeFileFS.getFileStatus(largeFile);
largeFileSize = fileStatus.getLen();
numBlocks = calculateNumBlocks(largeFileSize, BLOCK_SIZE);
}
Contributor
Author
There was a problem hiding this comment.
no test uses more than one file, does it?
correct, they use only 1 file
| smallFile = path("iTestS3APrefetchingLruEviction"); | ||
| ContractTestUtils.writeDataset(getFileSystem(), smallFile, data, data.length, 16, true); | ||
| blockSize = conf.getInt(PREFETCH_BLOCK_SIZE_KEY, PREFETCH_BLOCK_DEFAULT_SIZE); | ||
| smallFileFS = new S3AFileSystem(); |
Contributor
There was a problem hiding this comment.
can use use getFileSystem() here
|
🎊 +1 overall
This message was automatically generated. |
virajjasani
changed the title
Contributor
Author
|
|
|
🎊 +1 overall
This message was automatically generated. |
steveloughran
requested changes
Jul 20, 2023
Contributor
steveloughran
left a comment
steveloughran
left a comment
There was a problem hiding this comment.
really, really close, just need to move from fields to local vars
| private void createSmallFile() throws Exception { | ||
| byte[] data = ContractTestUtils.dataset(SMALL_FILE_SIZE, 'x', 26); | ||
| smallFile = path("iTestS3APrefetchingLruEviction"); | ||
| smallFile = methodPath(); |
Contributor
There was a problem hiding this comment.
this should be made a local variable; class field removed as well as teardown
| private void createLargeFile() throws Exception { | ||
| byte[] data = ContractTestUtils.dataset(S_1K * 72, 'x', 26); | ||
| largeFile = path(fileName); | ||
| largeFile = methodPath(); |
Contributor
There was a problem hiding this comment.
this should be made a local variable; class field removed as well as teardown
steveloughran
requested changes
Jul 20, 2023
Contributor
steveloughran
left a comment
steveloughran
left a comment
There was a problem hiding this comment.
ok, sorry, my fault, i annotated the wrong line,
It's the filesystem instances largeFileFS and smallFileFS which should be replaced with local vars -they are both just getFileSystem() references, and there's no need to close them in teardown as the superclass does its work there
|
🎊 +1 overall
This message was automatically generated. |
|
💔 -1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
Contributor
Author
|
us-west-2:
|
Contributor
Author
|
i plan to initiate few more followup sub-task Jira later: a) LRU contents unit testable with some refactor |
steveloughran
approved these changes
Jul 24, 2023
Contributor
steveloughran
left a comment
steveloughran
left a comment
There was a problem hiding this comment.
LGTM
+1
will merge
Contributor
|
merged to trunk; now I will get to rebase my unbuffer() patch atop it. Can you do the homework of a cherrypick and retest onto branch-3? thanks |
Contributor
Author
|
sure thing, will create 3.3 backport PR and test it. Thank you for merging the PR! |
jiajunmao
pushed a commit
to jiajunmao/hadoop-MLEC
that referenced
this pull request
Feb 6, 2024
Contributed by Viraj Jasani
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.
Jira: HADOOP-18805