SQL + Hadoop + low-latency = Yes, it is possible (opens in new tab)

(datasalt.com)

6 pointsivanprado13y ago2 comments

2 comments

This whole post is based on the assumption that "Splout SQL" is something "novel" and "necessary". To me these are not clear, though.

I have prototyped some Big Data systems in the past months where we use Sqoop for exporting Hadoop to proprietary SQL databases. So far, so good. I wonder what are the real advantages of using Splout SQL in such cases.

ivanpradoOP13y ago

Three things differentiate Splout SQL from using Sqoop for exporting to an existing SQL database: 1) Scalability: Relational databases rarely scales, or are too expensive for big volumes of data. They don't work well with Hadoop. 2) Update isolation: In Splout SQL, database updating never affects serving queries as it is performed in a Hadoop cluster. 3) Atomicity: Datasets are deployed atomically in Splout SQL. That avoids inconsistency problems that arises in RDMS when updating existing databases.

j / k navigate · click thread line to collapse

2 comments

bigeater13y ago

This whole post is based on the assumption that "Splout SQL" is something "novel" and "necessary". To me these are not clear, though.

ivanpradoOP13y ago

j / k navigate · click thread line to collapse