1) May i write somehow into hbase through spark-sql?
2) Does the SHC supports write operations (https://github.com/hortonworks-spark/shc) ?
it is great if here exists some java examples
3) I need to create a lot new columns in hbase and put information there, is it good idea work with hbase through spark-sql?
or i need to setup something else on top of hbase (phoenix, hive etc)
You can use JavaHBaseContext bundled in hbase-spark package here. It offers bulk loading capabilities.