### _Why are the changes needed?_
When a kyuubi service connections to multiple engine clusters, we can only explicitly specify relevant parameters after the URL, for example:
`beeline -u 'jdbc:hive2://127.0.0.1:10009/?kyuubi.engineEnv.FLINK_HOME=/opt/flink;kyuubi.ha.namespace=kyuubi-ns-c;kyuubi.engine.type=FLINK_SQL'`
This method is extremely unfriendly to users.
I implemented a SessionConfAdvisor implementation class `org.apache.kyuubi.session.FileSessionConfAdvisor` based on the conf file that to manage session level configurations, so that we can access this cluster only by specifying a unique engine name, for example:
access cluster-a:
`beeline -u 'jdbc:hive2://127.0.0.1:10009/?kyuubi.session.conf.profile=cluster-a'`
access cluster-c:
`beeline -u 'jdbc:hive2://127.0.0.1:10009/?kyuubi.session.conf.profile=cluster-c'`
`kyuubi-session-cluster-a.conf` configuration is as follows:
```
kyuubi.ha.namespace kyuubi-ns-a
kyuubi.engine.type SPARK_SQL
kyuubi.engine.pool.balance.policy POLLING
kyuubi.engineEnv.SPARK_HOME /opt/spark30
kyuubi.engineEnv.HADOOP_CONF_DIR /opt/hadoop_conf_dir
```
`kyuubi-session-cluster-c.conf` configuration is as follows:
```
kyuubi.ha.namespace kyuubi-ns-a
kyuubi.engine.type SPARK_SQL
kyuubi.engineEnv.SPARK_HOME /opt/spark32
kyuubi.engineEnv.HADOOP_CONF_DIR /opt/hadoop_conf_dir
```
### _How was this patch tested?_
- [x] Add some test cases that check the changes thoroughly including negative and positive cases if possible
- [ ] Add screenshots for manual tests if appropriate
- [x] [Run test](https://kyuubi.apache.org/docs/latest/develop_tools/testing.html#running-tests) locally before make a pull request
Closes#3742 from ychris78/yuqi_1101.
Closes#3742
e1449867 [yuqi] Style
68bbfecf [yuqi] restructure
1c8daa86 [yuqi] Change the cache expiration time to 10 minutes and remove the manual cache
5b4a6d77 [Cheng Pan] Update kyuubi-server/src/main/scala/org/apache/kyuubi/session/FileSessionConfAdvisor.scala
d332e2d8 [Cheng Pan] Update kyuubi-server/src/main/scala/org/apache/kyuubi/session/FileSessionConfAdvisor.scala
cb644ab4 [Cheng Pan] Update docs/deployment/settings.md
1ee3c980 [Cheng Pan] Update kyuubi-server/src/test/resources/kyuubi-session-cluster-a.conf
30f8fde0 [Cheng Pan] Update kyuubi-common/src/main/scala/org/apache/kyuubi/config/KyuubiConf.scala
5ef5841a [yuqi] Cache upgrade
5c2f7461 [yuqi] add cache and fix bug
1b1f9d67 [yuqi] scalastyle
a0507847 [yuqi] License info amendment
335250fc [yuqi] fix bug
9e8d313a [yuqi] Modify according to suggestions
5f57b637 [yuqi] Modify according to suggestions
5e504252 [yuqi] add org.apache.kyuubi.plugin.FileSessionConfAdvisor When kyuubi.session.conf.advisor=org.apache.kyuubi.plugin.FileSessionConfAdvisor,Specify a session level configuration file, which will be combined with default.conf to have an impact. The corresponding configuration file is conf/kyuubi-session-<profile>.conf
0cb71341 [yuqi] add org.apache.kyuubi.plugin.YamlSessionConfAdvisor
b524c221 [yuqi] add org.apache.kyuubi.plugin.YamlSessionConfAdvisor
51729d15 [yuqi] When a kyuubi service connections to multiple engine clusters, we can only explicitly specify relevant parameters after the URL, for example: beeline -u 'jdbc:hive2://127.0.0.1:10009/?kyuubi.engineEnv.FLINK_HOME=/opt/flink;kyuubi.ha.namespace=kyuubi-ns-c;kyuubi.engine.type=FLINK_SQL' This method is extremely unfriendly to users, and it is extremely unsafe to expose namespaces to users. So I added a yaml file(engine-cluster-env.yaml) to manage the engine clusters, so that we can use this cluster only by specifying a unique engine name, for example: use cluster-a: beeline -u 'jdbc:hive2://127.0.0.1:10009/?kyuubi.engine.cluster.name=cluster-a'
Lead-authored-by: yuqi <1450636878@qq.com>
Co-authored-by: Cheng Pan <pan3793@gmail.com>
Signed-off-by: Cheng Pan <chengpan@apache.org>
8788c3b2f3