[
https://issues.apache.org/jira/browse/DRILL-3886?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14941521#comment-14941521
]
Khurram Faraaz commented on DRILL-3886:
---------------------------------------
Data inserted into HBase table was in ASCENDING order,
org.apache.hadoop.hbase.util.OrderedBytes.encodeInt32(br, arr[i],
org.apache.hadoop.hbase.util.Order.ASCENDING);
Query from Drill was trying to get data in DESCENDING order by specifying
INT_OBD, and hence there were no results returned. Changing the encoding to
INT_OB query returned correct results.
> Range query over HBase table from Drill returns no results
> ----------------------------------------------------------
>
> Key: DRILL-3886
> URL: https://issues.apache.org/jira/browse/DRILL-3886
> Project: Apache Drill
> Issue Type: Bug
> Components: Execution - Flow
> Affects Versions: 1.2.0
> Environment: 4 node cluster CentOS
> Reporter: Khurram Faraaz
> Assignee: Smidth Panchamia
>
> Range query over HBase table from Drill returns no results. Note that the
> filter is pushed down to the Scan, however query does not return any results.
> Drill master commit ID: b9afcf8f
> case 1) without cast and without quotes around integer value
> {code}
> 0: jdbc:drill:schema=dfs.tmp> explain plan for select
> convert_from(row_key,'INT_OBD') as rk from int_Tbl WHERE
> convert_from(row_key,'INT_OBD') > -536870912 AND
> convert_from(row_key,'INT_OBD') < 536870911;
> +------+------+
> | text | json |
> +------+------+
> | 00-00 Screen
> 00-01 Project(rk=[CONVERT_FROMINT_OBD($0)])
> 00-02 Scan(groupscan=[HBaseGroupScan [HBaseScanSpec=HBaseScanSpec
> [tableName=int_Tbl, startRow=\xD4`\x00\x00\x00\x00,
> stopRow=\xD4\x9F\xFF\xFF\xFF, filter=FilterList AND (2/2): [RowFilter (LESS,
> \xD4\x9F\xFF\xFF\xFF), RowFilter (GREATER, \xD4`\x00\x00\x00)]],
> columns=[`row_key`]]])
> 0: jdbc:drill:schema=dfs.tmp> select convert_from(row_key,'INT_OBD') as rk
> from int_Tbl WHERE convert_from(row_key,'INT_OBD') > -536870912 AND
> convert_from(row_key,'INT_OBD') < 536870911;
> +--+
> | |
> +--+
> +--+
> No rows selected (0.869 seconds)
> {code}
> case 2) with cast and without quotes around integer value
> {code}
> 0: jdbc:drill:schema=dfs.tmp> select convert_from(row_key,'INT_OBD') as rk
> from int_Tbl WHERE convert_from(row_key,'INT_OBD') > cast(-536870912 as int)
> AND convert_from(row_key,'INT_OBD') < cast(536870911 as int);
> +--+
> | |
> +--+
> +--+
> No rows selected (0.905 seconds)
> 0: jdbc:drill:schema=dfs.tmp> explain plan for select
> convert_from(row_key,'INT_OBD') as rk from int_Tbl WHERE
> convert_from(row_key,'INT_OBD') > cast(-536870912 as int) AND
> convert_from(row_key,'INT_OBD') < cast(536870911 as int);
> +------+------+
> | text | json |
> +------+------+
> | 00-00 Screen
> 00-01 Project(rk=[CONVERT_FROMINT_OBD($0)])
> 00-02 Scan(groupscan=[HBaseGroupScan [HBaseScanSpec=HBaseScanSpec
> [tableName=int_Tbl, startRow=\xD4`\x00\x00\x00\x00,
> stopRow=\xD4\x9F\xFF\xFF\xFF, filter=FilterList AND (2/2): [RowFilter (LESS,
> \xD4\x9F\xFF\xFF\xFF), RowFilter (GREATER, \xD4`\x00\x00\x00)]],
> columns=[`row_key`]]])
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
