[
https://issues.apache.org/jira/browse/HBASE-15572?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15239584#comment-15239584
]
Jerry He commented on HBASE-15572:
----------------------------------
bq. For example, (select col1 from hbaseTable where rowkey = 1 and TIMESTAMP >
123345)
bq. However, it may cause potential troubles. For example, if a virtual column
is put into the sql schema, users may have select timestamp from table. It is a
valid query in sql, but not valid in Hbase, since this timestamp is implicit in
Hbase attached to a row, there is really no such column in the Hbase row. We
cannot construct a scan using this. Thus it is an invalid query for HBase.
Selecting the timestamp from table (have it in projection) is tricky, but it
could still be valid for advanced use case. Having it in predicates should
anyway work.
This support is for SQL. Shouldn't we leverage more of the flexibility of SQL?
But it is probably for another discussion.
The patch looks good.
> Adding optional timestamp semantics to HBase-Spark
> --------------------------------------------------
>
> Key: HBASE-15572
> URL: https://issues.apache.org/jira/browse/HBASE-15572
> Project: HBase
> Issue Type: New Feature
> Components: spark
> Reporter: Weiqing Yang
> Assignee: Weiqing Yang
> Fix For: 2.0.0
>
> Attachments: HBASE-15572-1.patch, HBASE-15572-10.patch,
> HBASE-15572-2.patch, HBASE-15572-3.patch, HBASE-15572-4.patch,
> HBASE-15572-5.patch, HBASE-15572-6.patch, HBASE-15572-7.patch,
> HBASE-15572-8.patch, HBASE-15572-9.patch
>
>
> Right now the timestamp is always latest. With this patch, users can select
> timestamps they want.
> In this patch, 4 parameters, "timestamp", "minTimestamp", "maxiTimestamp" and
> "maxVersions" are added to HBaseSparkConf. Users can select a timestamp, they
> can also select a time range with minimum timestamp and maximum timestamp. A
> new test for selecting records with different timestamps is added.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
