[
https://issues.apache.org/jira/browse/PHOENIX-2075?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14606464#comment-14606464
]
James Taylor commented on PHOENIX-2075:
---------------------------------------
Thanks, [~tdsilva]. Here's some feedback:
- Can we fix the problem with splitting up the column names and then
recombining them first, in a separate JIRA and then get rid of this change from
this JIRA?
{code}
diff --git
a/phoenix-core/src/main/java/org/apache/phoenix/mapreduce/util/ColumnInfoToStringEncoderDecoder.java
b/phoenix-core/src/main/java/org/apache/phoenix/mapreduce/util/ColumnInfoToStringEncoderDecoder.java
index ec52fba..c9dfa10 100644
---
a/phoenix-core/src/main/java/org/apache/phoenix/mapreduce/util/ColumnInfoToStringEncoderDecoder.java
+++
b/phoenix-core/src/main/java/org/apache/phoenix/mapreduce/util/ColumnInfoToStringEncoderDecoder.java
@@ -33,7 +33,7 @@ import com.google.common.collect.Lists;
*/
public class ColumnInfoToStringEncoderDecoder {
- private static final String COLUMN_INFO_DELIMITER = "|";
+ private static final String COLUMN_INFO_DELIMITER = "~";
private ColumnInfoToStringEncoderDecoder() {
{code}
- Add class level javadoc for ParallelScanGrouper and
DefaultParallelScanGrouper.
- How about making a static INSTANCE variable (or getInstance() static method)
in DefaultParallelScanGrouper and MapReduceParallelScanGrouper, make the
constructor private, and use the INSTANCE variable instead of constructing a
new one each time (as it's stateless)?
> MR integration uses single mapper unless table is salted
> --------------------------------------------------------
>
> Key: PHOENIX-2075
> URL: https://issues.apache.org/jira/browse/PHOENIX-2075
> Project: Phoenix
> Issue Type: Bug
> Reporter: James Taylor
> Assignee: Thomas D'Silva
> Attachments: PHOENIX-2075-4.4-HBase-0.98-v2.patch,
> PHOENIX-2075-4.4-HBase-0.98.patch
>
>
> Due to the way we interpret the scans, when we read data from our MR
> integration, it's always processed by a single mapper.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
