Code Monkey home page Code Monkey logo

Comments (2)

nik-io avatar nik-io commented on September 6, 2024

This might have to do with the amount of dynamic table grows required to fit the larger dataset. The default table size is equal to a single segment, which defaults to 512MiB, I believe

#define UDEPOT_DEFAULT_SEGMENT_SIZEB (1UL<<29) // bytes

You can try increasing that to 1 or 2 GiB and see if it helps.
Another thing that might affect performance could be the contention on the hash table. You can check if that’s the case if you run single thread mode and the performance issue goes away. In that case, you could increase the number of hash table locks.
If none of the above work, then I suggest to try different, growing combinations of thread count and trt-tasks parameter (per thread task parallelism) and see how they scale for the different test sizes. These should help identify where the problem is.

from udepot.

Cha465 avatar Cha465 commented on September 6, 2024

Thanks! Found out that load imbalance caused this (I have modified workload distribution to zipfian and used customized network stack).

from udepot.

Related Issues (3)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    πŸ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. πŸ“ŠπŸ“ˆπŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❀️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.