Can you recommend a high-performance, thread-safe and stable ORM for Python? The data I need to work with isn't complex, so SQLAlchemy is probably an overkill.
I had read that thread before; however, it doesn't really answer my question. I need something that can handle >5k queries per second with lots of writes, and I'm afraid it will take quite a lot of time to benchmark and test all the available ORMs, so I would gladly use any advice from those who have experience with that sort of thing.
David Parunakian
2010-08-31 09:21:27
A:
You can use a more declarative layer on top of SQLAlchemy such as Elixir, or look at Storm which is also a bit more declerative than SQLAlchemy.
The latter was developed for and is used by sites such as Ubuntu/Canonical's launchpad, so it should scale well.
Ivo van der Wijk
2010-08-31 09:41:03
A:
If you are looking for something thats high performance, and based on one of your comments "something that can handle >5k queries per second". You need to keep in mind that an ORM is not built specifically for speed and performance, it is built for maintainability and ease of use. If the data is so basic that even SqlAlchemy might be overkill, and your mostly doing writes, it might be easier to just do straight inserts and skip the ORM altogether.
Tanerax
2010-08-31 13:48:34
Thanks. I will also explore the possibility of switching to memcachedb for simple key-value pairs and fancy stuff like Cassandra for the rest if it will improve performance significantly without taking too much time to write.
David Parunakian
2010-09-01 13:44:09