Matthew Rocklin

And they have slightly better support :) On Thu, Jun 13, 2024 at 6:07 AM Matthew Rocklin ***@***.***> wrote: > Probably Actors are better for this kind of thing? >...

Tasks can take a long time (several hours is common) that is unrelated to the timeout (unless your task holds the GIL, which is uncommon unless you're wrapping custom C...

Lifetime was intended to provide a mechanism for a periodic reset of workers, such as might be useful when dealing with libraries that might leak memory and so benefit from...

I recommend updating dask and distributed. This looks like an issue that was recently fixed and released. On Sun, Jan 21, 2024 at 8:50 PM RichardScottOZ ***@***.***> wrote: > Attempted...

I have also experienced this anecdotally On Thu, Nov 17, 2022 at 10:40 AM James Bourbeau ***@***.***> wrote: > Recently we added a new button for launching a default set...

Hrm, I hadn't ever thought of that. Yes, I can see how that would suck. On Tue, Sep 12, 2023 at 12:35 PM crusaderky ***@***.***> wrote: > Use case >...

Here is a rendered notebook from my Ubuntu Thinkpad to make the numbers more concrete: https://gist.github.com/mrocklin/526120bb5231cc5d9d4e3ca87fc09c68 This run was actually a bit faster than I usually see. On Cloud VMs...

To be clear, my experiment in this issue was run without Dask and without Coiled. I wanted to isolate things.

Turning threads off I get around 170MB/s on my linux machine, 600 MB/s on my OS-X machine. > as I get around 250-300 MB/s using use_threads=False. I'm curious, for deserializing...

> But for linux you mentioned above a similar number with threads. So that means you see hardly any perf improvement with threads on linux Yes. That's correct. To be...