Re: One-shot semantics in GHC event manager

Hey Andreas, The basic rundown is that if we equip you with an account, you can just do it yourself. Although we'd like to restrict access a bit more; I'll figure something out. Yeah, if you hop on IRC, we can chat quickly about it and work something out in the mean time. On Tue, Oct 14, 2014 at 10:57 AM, Andreas Voellmy wrote:

This is awesome. I'd like to try to recreate some of the evaluations for the multicore IO manager paper on that 40 core system at backspace. How can I get access to this? I'll jump on IRC - maybe it is easier to chat in realtime.

On Mon, Oct 13, 2014 at 5:33 PM, Austin Seipp wrote:

...

For the record, I talked to Ben earlier on IRC, and I can provide him with a machine to do intense benchmarks of the new I/O manager.

Also, if any other developers (like Andreas, Johan, Bryan, etc) in this space want a big machine to test it on, I can probably equip you with one (or several). Since Rackspace is so gracious to us, we can immediately allocate high-powered, physical (i.e. not Xen, but real machines) machines to do high-scale testing on.[1]

In any case, it's not hard to do and only takes a few minutes, so Ben: let me know. (I've thought it would be neat to implement a leasing system somehow, where a developer could lease a few machines for a short period of time, at which point they expire and a background job cleans them up.)

[1] You can find the hardware specs here; GHC benchmarking is probably best suited for the "OnMetal I/O" type, which has 40 cores, 2x PCIe flash and 128GB of RAM - http://www.rackspace.com/cloud/servers/onmetal/

On Mon, Oct 13, 2014 at 2:05 PM, Ben Gamari wrote:

...

Ben Gamari writes:

...

Andreas Voellmy writes:

...

On Sat, Oct 11, 2014 at 12:17 PM, Ben Gamari wrote:

Ah... so this is not useful to you. I guess we could add `loop` to GHC.Event's export list. On the other hand, I like your LifeTime proposal better and then no one needs `loop`, so let's try this first.

I have a first cut of this here [1]. It compiles but would be I shocked if it ran. All of the pieces are there but I need to change EventLifetime to a more efficient encoding (there's no reason why it needs to be more than an Int).

As it turns out the patch seems to get through the testsuite after a few minor fixes.

What other tests can I subject this to? I'm afraid I don't have the access to any machine even close to the size of those that the original event manager was tested on so checking for performance regressions will be difficult.

Cheers,

- Ben

_______________________________________________ ghc-devs mailing list ghc-devs@haskell.org http://www.haskell.org/mailman/listinfo/ghc-devs

-- Regards,

Austin Seipp, Haskell Consultant Well-Typed LLP, http://www.well-typed.com/

-- Regards, Austin Seipp, Haskell Consultant Well-Typed LLP, http://www.well-typed.com/