Thrill – Big Data Processing with C++

http://blog.brakmic.com/thrill-big-data-processing-with-c/

by brakmic • 9 years ago

127 21 9 years ago

blog.brakmic.com

pzh2• 9 years ago

Does anybody know how this is different from Spark? These Distributed Immutable Arrays sound suspiciously similar to Spark's Resilient Distributed Datasets. Is it just the choice of C++ as opposed to Scala that would make this more efficient?

Also, I wonder if and how they implemented the concept of lineage (unless these DIAs are not really very resilient)... I thought Spark relied on Scala's delayed evaluation to do that, though I may be mistaken.

beached_whale• 9 years ago

In the presentation they directly contrast the performance of the two. http://panthema.net/2016/1206-Thrill-High-Performance-Algori...

sreenadh• 9 years ago

Is there a video on the talk?

lorenzhs8• 9 years ago

DIAs are quite heavily inspired by RDDs. A lot of the performance increases come from the C++ compiler's ability to fuse local operations etc.

Thrill doesn't implement any fault tolerance at the moment, it's closer to prototype status than production readiness.

pzh2• 9 years ago

Do you have any plans regarding how you'd implement resilience and the equivalent of Spark's concept of 'lineage', where you keep a history of how a given RDD was computed, and then you can recompute it if it gets lost?

I haven't looked into Spark in depth, but I believe that 'lineage' relies heavily on Scala's delayed evaluation and the underlying Java RMI facilities. Doing something similar in C++ may require a lot more effort and a significantly different set of tradeoffs regarding the performance model.

willtim2• 9 years ago

The JVM was obviously never designed for functional programming. I would like to see something like Thrill built in Haskell or OCaml, both of which generate efficient native code, support unboxed arrays and have metaprogramming/staging facilities that go beyond templates. GHC Haskell even has declarative rewrite rules for compile-time fusion.

I'm still not convinced C++ is necessary to outperform Spark, especially as high-level features like reference counting and lambdas are being used.

Mikeb85• 9 years ago

Very cool. Will have to remember this, maybe write an R package that makes use of it.

fnord123• 9 years ago

Not using thrill, but there is Rpbd:

https://rbigdata.github.io/

It's basically R on scalapack.

tmsldd• 9 years ago

The Force is strong with this one

codepie• 9 years ago

There's also Blogel [0] which is a distributed graph processing framework in C++ and it runs significantly faster than its counterpart in Java, Apache Giraph [1].

I have started wondering if the big data developers really care about the speed; the advantages of these Java softwares start to fade out when compared with their C++ counterparts.

[0] http://www.cse.cuhk.edu.hk/blogel/

[1] http://www.cse.cuhk.edu.hk/blogel/papers/blogel.pdf

pjmlp• 9 years ago

If you just measure milliseconds yes.

If you measure project costs, including the salaries of the developers and amount of development days, then no.

This is the main reason why there is such a big pressure from trading folks for Oracle to improve Java regarding value types and FFI to native code.

lorenzhs8• 9 years ago

I think with Thrill, there are two different skill levels to be distinguished:

- Using it to implement things should be fairly easy and doesn't require advanced knowledge of C++. Basically you have to plug lambdas that do the processing into the provided operations, similar to Spark, but using C++ syntax. It might require some compiler error parsing skills, but altogether it shouldn't be too different from using Spark with Java/Scala

- Extending Thrill requires familiarity with modern C++, possibly including advanced template tricks.

Since there isn't a whole lot of advanced stuff available for Thrill (yet), that means that currently people with the latter skills would most likely be required at the moment. But in a world where the same libraries available for Spark are available for Thrill or a similar C++ framework, that wouldn't be the case. Note that Thrill is currently quite experimental.

I guess it's a trade-off, but dismissing the potential for 10x runtime gains "because C++" seems too one-sided. That isn't to say that the C++ frameworks don't have a long way to go before they can rival Spark etc in ease of use and tooling, they do! But at least they point out the inefficiencies and potential for improvement in these existing systems.

adrianN• 9 years ago

There is also the STXXL [1] for times when your data is big but not "big". It contains containers and algorithms optimized for external storage.

http://stxxl.sourceforge.net/

lorenzhs8• 9 years ago

Thrill and STXXL are both delevoped in the same group at KIT (I work there, too, but I'm not directly involved). Thrill also reuses some parts of STXXL, and does so completely transparently to the user - if memory doesn't suffice, it'll use the disk.

lorenzhs8• 9 years ago

The companion paper to Thrill, with more details on its architecture and some benchmarks and comparisons to Spark and Flink: https://arxiv.org/abs/1608.05634

75dvtwin• 9 years ago

Thx for the link to the paper. It is a useful read in its own right.

It provides an overview of other processing framework, explains why C++ was chosen, explains various bottlenecks and their affect.

I had worked with KMeans before, and happy to see as part of bench-marking, as it seems one of the more widely used approaches for unsupervised learning.

In my view, Thrill is similar in composite-ability and integration into existing code objectives to Python's Dask

Thrill – Big Data Processing with C++

Related Stories