|
|
|
|
|
|
by zobzu
5334 days ago
|
|
|
I've had similar performance in my use case (big joins and very large tables) using PostgreSQL (in my case) and disabling sync() to disk, and tuning the buffers, as with the various NoSQL I tried. It seems to me that NoSQL does not really bring speed. Just scalability and a different model. Hopefully most of them don't lose data at random. PG certainly doesn't, even with sync() off. I have not tested the scalability of PG. |
|
|
Aren't joins one of the things that NoSQL loudly and proudly announces that it isn't suitable for?