Asynchronous I/O in DuckDB: Work, Thread, Work

(duckdb.org)

181 points | by pdet 5 days ago

10 comments

  • diarrhea 3 hours ago
    Does having the worker pool hold as many threads as cores work well alongside the async pool? It is basically oversubscribed by design.

    I built a system once which had (this is Rust) a Rayon worker thread pool of 4 threads and a Tokio async pool of 2 (multithreaded runtime). On a system of 6 vCPU. This ended up working fine. Tokio was not starved so handled network requests at low latency.

    One difference is DuckDB is a pure network client. If one of its async threads is starved it is not the end of the world (e.g. k8s does not kill your pod for failure of replying to health checks).

    • Asmod4n 12 minutes ago
      Network IO is heavily NIC queue bound, if your NIC only has one queue it just makes it slower to do any threading workload against it.

      On my ryzen 9 it needs around 8 cores to do the same work in a threaded io loop than you can do single threaded. And the mechanism doesnt matter, you could share an fd, use SO_REUSEPORT or just share memory between threads.

      Just doing the sharing makes everything extremely slow. One context switch becomes more expensive than just doing it single threaded.

    • pdet 2 hours ago
      I've run quite a few benchmarks on that as well, on a few different machines, and oversubscribing ASYNC threads demonstrated very little performance downside. In the end, the memory governor also keeps these threads "in check" while still allowing full utilization when possible.

      There is still something to gain from tuning it further (as you can see in the async I/O tuned benchmark), but having that network saturation by default is still a work in progress.

      (Disclaimer: I'm the author of the blogpost)

    • marginalia_nu 2 hours ago
      As long as you're scheduled by the kernel and not something like Kubernetes with a CPU limit, you can generally oversubscribe I/O threads without much of a problem. They're mostly parked waiting for syscalls anyway. Heck, even if they're mostly doing CPU work, the scheduler generally deals with it pretty gracefully.
  • bburnett44 7 hours ago
    Using 512gb of ram for a 22gb remote file does feel a bit weird for a benchmark but maybe they couldn’t get a large number of cores without lots of memory?
    • pdet 2 hours ago
      The main reason I decided to use a beefier machine is that it gives me flexibility when benchmarking, without the need to set up different environments. The CSV data, for example, is >80 GB. We can also “scale down the machine” for experiments where we want to stress-test lower-memory scenarios or use fewer threads by configuring DuckDB’s settings (e.g., SET memory_limit = '10GB'; or SET threads TO 1;).

      (Disclaimer: I’m the author of the blog post.)

    • otterley 7 hours ago
      Most cloud providers start with a 2:1 ratio of memory in GiB to CPU cores and go up from there. Databases also are the most common workload for large-memory systems because they benefit so much from large buffer caches.
    • marginalia_nu 3 hours ago
      Even so you could use cgroups to cap the memory for the process (including its caches).
  • ilyagr 1 hour ago
    Do the CSV files allow quoted newlines? If yes, what's the trick to avoid checking the whole file too find out whether a newline is quoted or a record separator when reading it from the middle in an async thread?
    • pdet 1 hour ago
      DuckDB uses a speculative parallel CSV parsing technique. The basic idea is that the parser speculates about the state the CSV parser is in at a random byte (e.g., whether it is inside a quoted field) and tries to figure out where the next row starts based on that.There are validation steps during finalization as well, to ensure the parser did not got anything wrong in its speculation.

      I've never gotten around to writing a blog post about it, but I go quite in-depth on the technique in this presentation: https://www.youtube.com/watch?v=YrqSp8m7fmk

      (Disclaimer: I'm the author of the blog post and also the developer who implemented the entire CSV parser in DuckDB.)

      • hilariously 35 minutes ago
        Burning my points to say dude that's sick, parsing CSVs is hell I am genuinely going to watch this thank you.
      • abirch 1 hour ago
        Thank you so much. I love DuckDB!
    • xxs 1 hour ago
      most likely - they read large chunks and decode them
  • myshapeprotocol 9 hours ago
    Deep dive into asynchronous I/O architectures like this is pure engineering gold for high-performance data processing. Excellent breakdown.
    • xxs 1 hour ago
      I can't say it's 'deep' in any way, e.g. what type of queue is that? Another part - utmost importance to keep async threads fully busy - that doesn't address what kind of disk subsystem is and how it deals with random reads.
  • datadrivenangel 7 hours ago
    DuckDB is trending towards becoming a query engine, specifically the fastest analytical query engine. This is very good.
  • pjot 6 hours ago
    I wonder how this would work in trying to parallelize the worker threads (multiple duckdb instances) coordinating them via Quack.

    Ducks all the way down!

  • abofh 7 hours ago
    Do they have SSL updates yet? Signing is great, but using https means not fighting firewalls to start a job
  • singhutsav5501 3 hours ago
    is there any aggregator for official docs such as this for db/systems/distributed arch at scale?
  • mansi1010 9 hours ago
    This is such a long waited feature!
  • myshapeprotocol 5 hours ago
    Brilliant deep dive into asynchronous I/O and execution thread models. Essential reading for high-performance data engineering.