We developed the solution based on Hadoop, but using Hadoop as a relationship database, it's hard to maintain and recruit developers. There needs to be better integrations with other solutions and there need to be more tools (including virtualization tools) and utilities between this solution and Hadoop.

For example, Oracle has a product called Big Data SQL and it built a bridge between Hadoop and Oracle's database to make things easier for developers.

The solution needs to be more stable.

The downloading process should be improved.

The solution needs to improve the functionality of the auto-deduplication. 

We should be able to automatically archive, and right now we can't.