[
https://issues.apache.org/jira/browse/FLINK-21782?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Felipe Lolas updated FLINK-21782:
---------------------------------
Description:
Hi,
Currently in a HBase table we model the data for a fast prefix filter scan,
would be nice that we can use a lookup join as:
{code:java}
SELECT
*
FROM
tbl
LEFT JOIN
hbaseTbl
ON
hbaseTbl.rowkey between tbl.field1 AND tbl.field2{code}
In HBase we can traduce that as a RANGE SCAN operator with startRow =
tbl.field1 and endRow = tbl.field2
Maybe too (same as Apache phoenix behaviour)
{code:java}
SELECT
*
FROM
tbl
LEFT JOIN
hbaseTbl
ON
hbaseTbl.rowkey LIKE tbl.field1 || '%'{code}
traduced as RANGE SCAN with startRow=tbl.field1 - 1 (replacing the last
character with lexicographical order -1) and endRow=tbl.field1 or simply a
RowPrefixFilter (does the same)
was:
Hi,
Currently in a HBase table we model the data for a fast prefix filter scan,
would be nice that we can use a lookup join as:
{code:java}
SELECT
*
FROM
tbl
LEFT JOIN
hbaseTbl
ON
hbaseTbl.rowkey between tbl.field1 AND tbl.field2{code}
In HBase we can traduce that as a RANGE SCAN operator with startRow =
tbl.field1 and endRow = tbl.field2
Maybe too (same as Apache phoenix behaviour)
{code:java}
SELECT
*
FROM
tbl
LEFT JOIN
hbaseTbl
ON
hbaseTbl.rowkey LIKE tbl.field1 || '%'{code}
traduced as RANGE SCAN with startRow=tbl.field1 - 1 (replacing the last
character with lexicographical order -1) and endRow=tbl.field1
> Lookup Table based on rowkey ranges
> -----------------------------------
>
> Key: FLINK-21782
> URL: https://issues.apache.org/jira/browse/FLINK-21782
> Project: Flink
> Issue Type: Improvement
> Reporter: Felipe Lolas
> Priority: Major
>
> Hi,
> Currently in a HBase table we model the data for a fast prefix filter scan,
> would be nice that we can use a lookup join as:
>
> {code:java}
> SELECT
> *
> FROM
> tbl
> LEFT JOIN
> hbaseTbl
> ON
> hbaseTbl.rowkey between tbl.field1 AND tbl.field2{code}
> In HBase we can traduce that as a RANGE SCAN operator with startRow =
> tbl.field1 and endRow = tbl.field2
> Maybe too (same as Apache phoenix behaviour)
>
> {code:java}
> SELECT
> *
> FROM
> tbl
> LEFT JOIN
> hbaseTbl
> ON
> hbaseTbl.rowkey LIKE tbl.field1 || '%'{code}
>
> traduced as RANGE SCAN with startRow=tbl.field1 - 1 (replacing the last
> character with lexicographical order -1) and endRow=tbl.field1 or simply a
> RowPrefixFilter (does the same)
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
