[
https://issues.apache.org/jira/browse/HBASE-27705?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17750522#comment-17750522
]
Zhen Wang commented on HBASE-27705:
-----------------------------------
We have multiple HBase clusters independent of Spark's computing cluster, and
we provide an Apache Kyuubi service for submitting Spark SQL jobs which do not
contain the `hbase-site.xml` configuration files.
Since Spark can use the configuration beginning with `spark.hadoop.` to add
hadoop configurations, so I want to use `spark.hadoop.hbase.XXX` to add hbase
configurations, like:
```
spark.hadoop.hbase.zookeeper.quorum=XXX;
spark.hadoop.hbase.security.authentication=kerberos;
```
By the way, for Spark to access hive's hbase external table through
`hive-hbase-handler`, we can also use `spark.hadoop.hbase.XXX` to specify hbase
configurations.
> Respect SparkContext hadoop configuration
> -----------------------------------------
>
> Key: HBASE-27705
> URL: https://issues.apache.org/jira/browse/HBASE-27705
> Project: HBase
> Issue Type: Improvement
> Components: spark
> Environment: spark: 3.1.1
> hbase-connector: 1.0.1-SNAPSHOT
> Reporter: Zhen Wang
> Priority: Major
> Fix For: hbase-connectors-1.0.1
>
>
> In our environment, the hbase clusters are independent of spark computing
> cluster, I want to set hbase configuration through `spark.hadoop.hbase.XXX`,
> so I hope hbase-spark can respect `sparkContext.hadoopConfiguration`.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
