4 ms·

I'm glad to see the paper specifically contrast iterator approaches (pull model) versus a "data-centric" approach (push model). This was indeed the issue I've s

I'm glad to see the paper specifically contrast iterator approaches (pull model) versus a "data-centric" approach (push model). This was indeed the issue I've seen with Apache Spark's RDD. Imagine a common pipeline: - input data - process data - output data Using iterators, the output logic loops over an abstract next() interface provided by the process logic, which in turn loops over another abstract next() interface provided by the input logic. Given a long enough chain of dependent processes, the CPU's cache will be obliterated. If we were to write code in a regular programming language by hand, we would never use such an inverted model. Instead, we would directly input the data, process it, and then output it. That's exactly what the paper proposes; the authors supply a callback function at the major stages of query evaluation. They use this model in the context of writing a compiler to speed query execution. It's worth noting that the Spark maintainers also realized that iterators have bad performance and that newer versions of Spark follow the compiler approach: https://databricks.com/blog/2016/05/23/apache-spark-as-a-compiler-joining-a-billion-rows-per-second-on-a-laptop.html https://databricks.com/blog/2016/05/23/apache-spark-as-a-com...

atombender 8y ago
As someone working on exactly such a query engine, this is very interesting, thanks. It's worth noting that Rust uses iterators extensively, yet is able to get around much of the issues with the pull model by being able to optimize and inline iteration loops. I believe its starting point for getting good performance here is by making iterators — i.e., the std::iter::Iterator trait — a first-class trait. For example, "for v in iter" knows about how to loop over iterators, and "for v in values" sees if "values" implements the IntoIter trait to turn it into an iterator. Rust also avoids vtable lookup when the value is not a polymorphic boxed trait. Together, this means it can aggressively inline and unroll iteration loops. I'm sure they use other tricks, as well.
voidmain 8y ago
Not all kinds of query plans can be expressed as nested loops, though. Index intersections and other ordered merge operations are very useful, and how do you do them with only nested loop plans?

atombender 8y ago

As someone working on exactly such a query engine, this is very interesting, thanks. It's worth noting that Rust uses iterators extensively, yet is able to get around much of the issues with the pull model by being able to optimize and inline iteration loops. I believe its starting point for getting good performance here is by making iterators — i.e., the std::iter::Iterator trait — a first-class trait. For example, "for v in iter" knows about how to loop over iterators, and "for v in values" sees if "values" implements the IntoIter trait to turn it into an iterator. Rust also avoids vtable lookup when the value is not a polymorphic boxed trait. Together, this means it can aggressively inline and unroll iteration loops. I'm sure they use other tricks, as well.

voidmain 8y ago

Not all kinds of query plans can be expressed as nested loops, though. Index intersections and other ordered merge operations are very useful, and how do you do them with only nested loop plans?