BASIC CHALLENGES WITH OTHER SQL SOLUTIONS
- Other SQL solutions rely on the traditional process of manually creating schemas or metadata definitions in a centralized store.
Expensive ETL (extract, transform, and load) routines need to be performed upfront in order to transform the data into a format the SQL engine can ingest. Data consumers need to wait before data can be made available to them. Data owners need to double the storage footprint in order to store data in its original format and in the ingested format.
This means big data analytics has to slow down to wait for long IT cycles, limiting the opportunity for end users to quickly explore new datasets or make real-time decisions, effectively diminishing the power of big data analytics itself.
- Other SQL solutions rely on data to be ingested into a managed and proprietary storage format that's very weak in handling complex data types and rapidly changing data formats. Proprietary storage formats also means there is vendor lock-in.
- Other SQL solutions lack the ability to dynamically size compute clusters independent of storage clusters.
- For other SQL solutions, switching to highly available is manual.
- Other solutions lack an integrated web-based UI with which all common developer and administrator tasks can be carried out.