[
https://issues.apache.org/jira/browse/HBASE-10534?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13906604#comment-13906604
]
rajeshbabu commented on HBASE-10534:
------------------------------------
[~lhofhansl]
This issue related to HBASE-8768(Improve bulkload performance) which is not
available in 0.94.
Can we backport HBASE-8768 to 0.94 as well?
Thanks.
> Rowkey in TsvImporterTextMapper initializing with wrong length
> --------------------------------------------------------------
>
> Key: HBASE-10534
> URL: https://issues.apache.org/jira/browse/HBASE-10534
> Project: HBase
> Issue Type: Bug
> Components: mapreduce
> Affects Versions: 0.96.0
> Reporter: rajeshbabu
> Assignee: rajeshbabu
> Fix For: 0.96.2, 0.98.1, 0.99.0
>
> Attachments: HBASE-10534.patch
>
>
> In TsvImporterTextMapper rowkey initializing with wrong length.
> parser.parseRowKey give pair of start and end positions, so the rowkey length
> initialized with rowkey end position instead of length.
> {code}
> Pair<Integer,Integer> rowKeyOffests =
> parser.parseRowKey(value.getBytes(), value.getLength());
> ImmutableBytesWritable rowKey = new ImmutableBytesWritable(
> value.getBytes(), rowKeyOffests.getFirst(),
> rowKeyOffests.getSecond());
> {code}
> Its better to change TsvParser#parseRowKey to return starting position and
> length of rowkey.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
