Geek asks and answers

1 TB reporting / analytics database with a simple model

Big data = 1 TB, increasing by 10% every year. The model is simple .. one table with 25 columns. No joins to other tables.

I want to do a simple filtering of queries on a subset of 25 columns.

I would suggest that a traditional SQL repository with indexes in filtered columns is necessary. Hadoop is redundant and does not make sense since it is for real-time service. Mongo bi-engine like a pentaho?

Any recommendations?

+3
source share
3 answers

, , , .

NoSQL BI/Reporting.

. . . partitioning PostgreSQL .

P.S. ORACLE/DB2, PostgreSQL.

+3

Infobright - , TB, , , ( , OSS, - , ). , , , MySQL, . , , 50 . , Infobright Pentaho, .

- , no-admin - .

+1

It looks like column storage will help. depends on how you handle the inserts, and if you ever need to do updates. But just like indbright, if you are going to advertise, then put a bill in a vector, it is faster and similar.

If you want free / open source, look at Luciddb - There aren’t many documents there, but it’s very good at what it does!

If you need incredible speed, check out the vector pointer. I believe that the price is the same as the inbright, but much faster.

0
source

More articles:


All Articles