[
https://issues.apache.org/jira/browse/HBASE-16890?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15595654#comment-15595654
]
ramkrishna.s.vasudevan commented on HBASE-16890:
------------------------------------------------
This comment in FSDataOutputStreamer
{code}
// We want this value to be a multiple of 3 because the native code checksums
46 // 3 chunks simultaneously. The chosen value of 9 strikes a balance between
47 // limiting the number of JNI calls and flushing to the underlying stream
48 // relatively frequently.
49 private static final int BUFFER_NUM_CHUNKS = 9;
{code}
Where the byte[] and checksum[] are multiples of 9. Does that have an impact
here?
> Analyze the performance of AsyncWAL and fix the same
> ----------------------------------------------------
>
> Key: HBASE-16890
> URL: https://issues.apache.org/jira/browse/HBASE-16890
> Project: HBase
> Issue Type: Sub-task
> Components: wal
> Affects Versions: 2.0.0
> Reporter: ramkrishna.s.vasudevan
> Assignee: ramkrishna.s.vasudevan
> Fix For: 2.0.0
>
> Attachments: contention.png, contention_defaultWAL.png
>
>
> Tests reveal that AsyncWAL under load in single node cluster performs slower
> than the Default WAL. This task is to analyze and see if we could fix it.
> See some discussions in the tail of JIRA HBASE-15536.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
