[
https://issues.apache.org/jira/browse/CASSANDRA-15215?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17453616#comment-17453616
]
Aleksandr Sorokoumov commented on CASSANDRA-15215:
--------------------------------------------------
The issue was caused by the slow path in
{{BufferedDataOutputStreamPlus#writeBytes}} when the underlying buffer has less
than 8 bytes remaining. Previously, this method fell back to {{writeSlow}}.
This was not correct because it writes N least significant bytes to the wire.
As {{writeBytes}} treats the register as an optimized version of a byte array,
it should write N most significant bytes instead. I added a test case that
isolates the issue and fixed it in all branches.
[~benedict] can you please re-run the CI?
> VIntCoding should read and write more efficiently
> -------------------------------------------------
>
> Key: CASSANDRA-15215
> URL: https://issues.apache.org/jira/browse/CASSANDRA-15215
> Project: Cassandra
> Issue Type: Bug
> Components: Local/Compaction, Local/SSTable
> Reporter: Benedict Elliott Smith
> Assignee: Aleksandr Sorokoumov
> Priority: Normal
> Fix For: 3.0.x, 3.11.x, 4.x
>
> Attachments: testWriteRandomLongDOP_final.png,
> writeUnsignedVInt_megamorphic_BB.png, writeUnsignedVInt_megamorphic_DOP.png
>
> Time Spent: 40m
> Remaining Estimate: 0h
>
> Most vints occupy significantly fewer than 8 bytes, and most buffers have >=
> 8 bytes spare, in which case we can construct the relevant bytes in a
> register and memcpy them to the correct position. Since we read and write a
> lot of vints, this waste is probably measurable, particularly during
> compaction and flush, and can probably be considered a performance bug.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
