1. Before the tenant deletes, you can check whether the tenant and the resource have an association relationship. If there is any, delete the error and prompt the related item.
2. The workflow is the same as above.
3. The tenant code can be considered to adjust the restriction rules at the front end, allowing the encoding of pure numbers or underscores.
4. The data source password is encrypted, such as the data source test connection log, the workflow run log, the data source, and the password information in the connection parameters in the metadata table to improve system security.
5. If you can support hplsql better, you can execute the stored procedure on similar relational data on hive/impala/other; this is for users with special needs, such as the user wants to open a new table. Each record of the orc table or kudu table is processed, the processing rules are different, and the cursor in the stored procedure can satisfy similar functions.

1.租户删除前可以考虑校验租户和资源是否存在关联关系，如果有，删除时候，报错提示出来关联项。
2.工作流同上。
3.租户编码可以考虑在前端把限制规则调整一下，允许纯数字或者下划线这种的编码。
4.涉及数据源密码的地方需要加密，如数据源测试连接日志、工作流运行日志、查看数据源、元数据表中连接参数中密码信息，以提高系统安全性。
5.如果可以支持hplsql更好，就可以在hive/impala/other上执行类似关系型数据上的存储过程；这样对于有特殊需求的用户来说，比如用户想对某个表（开启更新的hive orc表或者kudu表）的每一条记录进行处理，处理规则是不同的，存储过程里的游标就可以满足类似功能。