Optimize GCS Mock#51593
Merged
original-brownbear merged 1 commit intoelastic:masterfrom Jan 29, 2020
original-brownbear:more-efficient-gcs-mock-master
Merged
Optimize GCS Mock#51593original-brownbear merged 1 commit intoelastic:masterfrom original-brownbear:more-efficient-gcs-mock-master
original-brownbear merged 1 commit intoelastic:masterfrom
original-brownbear:more-efficient-gcs-mock-master
Conversation
This test was still very GC heavy in Java 8 runs in particular which seems to slow down request processing to the point of timeouts in some runs. This PR completely removes the large number of O(MB) `byte[]` allocations that were happening in the mock http handler which cuts the allocation rate by about a factor of 5 in my local testing for the GC heavy `testSnapshotWithLargeSegmentFiles` run. Closes #51446
Collaborator
|
Pinging @elastic/es-distributed (:Distributed/Snapshot/Restore) |
Contributor
Author
|
@tlrx if this doesn't help I think we have to find a way to down-scale these tests a little maybe. They're handling multiple MB of data at once which is a little tricky with |
| final int start = getContentRangeStart(range); | ||
| final int end = getContentRangeEnd(range); | ||
|
|
||
| final ByteArrayOutputStream out = new ByteArrayOutputStream() { |
Contributor
Author
There was a problem hiding this comment.
Using the BAOS here was pretty brutal for multi MB requests :) Much nicer now with the 16kb chunks Streams.readFully(wrappedRequest) produces
| final byte[] buffer = new byte[len]; | ||
| Streams.readFully(stream, buffer); | ||
| content = Tuple.tuple(name, new BytesArray(buffer)); | ||
| content = Tuple.tuple(name, fullRequestBody.slice(startPos, len)); |
Contributor
Author
There was a problem hiding this comment.
Even though slice wastes some memory here for the wrapping around the actual blob content for smaller blobs, it's a huge efficiency gain to stick with 16kb chunks from the BytesReference compared to allocating a potentially MB sized array.
tlrx
approved these changes
Jan 29, 2020
Member
tlrx
left a comment
There was a problem hiding this comment.
LGTM, thanks for taking care of this Armin
| } | ||
| System.arraycopy(out.toByteArray(), 0, blob, start, Math.toIntExact(bytesRead)); | ||
| blobs.put(blobName, new BytesArray(blob)); | ||
| blob = new CompositeBytesReference(blob, Streams.readFully(wrappedRequest)); |
Contributor
Author
|
Thanks Tanguy! |
This was referenced Jan 29, 2020
original-brownbear
added a commit
that referenced
this pull request
Jan 29, 2020
This test was still very GC heavy in Java 8 runs in particular which seems to slow down request processing to the point of timeouts in some runs. This PR completely removes the large number of O(MB) `byte[]` allocations that were happening in the mock http handler which cuts the allocation rate by about a factor of 5 in my local testing for the GC heavy `testSnapshotWithLargeSegmentFiles` run. Closes #51446 Closes #50754
original-brownbear
added a commit
that referenced
this pull request
Jan 29, 2020
This test was still very GC heavy in Java 8 runs in particular which seems to slow down request processing to the point of timeouts in some runs. This PR completely removes the large number of O(MB) `byte[]` allocations that were happening in the mock http handler which cuts the allocation rate by about a factor of 5 in my local testing for the GC heavy `testSnapshotWithLargeSegmentFiles` run. Closes #51446 Closes #50754
original-brownbear
added a commit
that referenced
this pull request
Jan 29, 2020
For small uploads (that can still be up to 5MB!) we needlessly reading the `InputStream` into a BAOS which entailed allocating the `byte[]` for the stream contents twice (because to `toByteArray` on the BAOS copies). Also, for resumeable uploads we were needlessly wrapping the output channel and running each individual write in its own privileged context when we could just wrap the whole upload in a single privileged context. Relates #51593
original-brownbear
added a commit
that referenced
this pull request
Jan 29, 2020
For small uploads (that can still be up to 5MB!) we needlessly reading the `InputStream` into a BAOS which entailed allocating the `byte[]` for the stream contents twice (because to `toByteArray` on the BAOS copies). Also, for resumeable uploads we were needlessly wrapping the output channel and running each individual write in its own privileged context when we could just wrap the whole upload in a single privileged context. Relates #51593
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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.
This test was still very GC heavy in Java 8 runs in particular
which seems to slow down request processing to the point of timeouts
in some runs.
This PR completely removes the large number of O(MB)
byte[]allocationsthat were happening in the mock http handler which cuts the allocation rate
by about a factor of 5 in my local testing for the GC heavy
testSnapshotWithLargeSegmentFilesrun.
Closes #51446
Closes #50754