4 ms·

I would also recommend this article to anyone who struggles with making many threads do useful amounts of work: https://lmax-exchange.github.io/disruptor/disru

I would also recommend this article to anyone who struggles with making many threads do useful amounts of work: https://lmax-exchange.github.io/disruptor/disruptor.html https://lmax-exchange.github.io/disruptor/disruptor.html

anonymousDan 4y ago
In a similar vein, this is also an interesting approach to avoiding synchronisation overhead that works for many data structures: https://www.cs.bgu.ac.il/~hendlerd/papers/flat-combining.pdf https://www.cs.bgu.ac.il/~hendlerd/papers/flat-combining.pdf
- bob1029 4y ago
  This sounds very similar indeed. The central thesis of the paper seems to be that the grain of locking is the concern, which is what Disruptor directly touches on as well. In a Disruptor implementation, you are typically processing batches of things, and locks can be acquired along lines of batches.
- gpderetta 4y ago
  That's very interesting, thanks.