A few thoughts to consider when migrating traditional SQL-based enterprise information centers to the spark architecture ... [In the amendment]

A few things to consider when migrating traditional SQL-based enterprise information centers to the spark architecture

* Reason: fashion, this is not big enough?

> data is designed as No-sql mode, and only search is required to establish a Level 2 index. It is possible that the RDBMS structure is not required.

Search, and reports can be query with spark SQL.

And Spark provides built-in Mlib and GRAPHX General reports are enough.

> DBAs usually do a backup or something. How do I back up the data after switching to the spark system?

> Is there a feature for offsite backup?

A few thoughts to consider when migrating traditional SQL-based enterprise information centers to the spark architecture ... [In the amendment]