On 17 Jul 2017 06:49, "Dennis Raddle" wrote: Thanks. I guess I don't know much about cloud services. Will they run a lot of threads in parallel? Or do they run my program on only one CPU and/or use only the number of cores in one processor? That's the point of using them, to run more threads in parallel than I have cores on my MacBook. Depending on your budget, on AWS you can rent a machine from the miniscule t2.nano all the way up to the 128-core x1.32xlarge, with various tradeoffs on memory vs CPU vs storage vs ... in between. All can run Haskell programs. There's also various options in Haskell for distributing your computation across multiple machines, if that sounds useful. A good place to start is the pricing page which gives you a good overview of the options. https://aws.amazon.com/ec2/pricing/on-demand/ Other cloud providers are available, of course. Cheers, D _______________________________________________ Haskell-Cafe mailing list To (un)subscribe, modify options or view archives go to: http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe Only members subscribed via the mailman list are allowed to post.

check these folks too: https://cloud.google.com/products/calculator/ ​ pretty good choice

On Jul 18, 2017 22:38, "Dennis Raddle" wrote: Neat, I like that they have a Haskell environment. But, if I go with Google or Amazon, just curious about how I would create an app with my Mac, as they don't provide a macOS environment. Can I cross-compile for Windows or Linux on my Mac? Or, should I run a Windows virtual machine and build my app in that environment? (Not my favorite choice, as I would end up working in Windows most of the time.) I'd use docker and run a docker container. Stack has nice support for building them. D On Sun, Jul 16, 2017 at 11:07 PM, Lyndon Maydwell wrote:

I heard about https://www.clever-cloud.com/ providing Haskell app support.

Didn't get to try it out though.

Warm Regards,

- Lyndon

...

On 17 Jul 2017 06:49, "Dennis Raddle" wrote:

Thanks. I guess I don't know much about cloud services. Will they run a lot of threads in parallel? Or do they run my program on only one CPU and/or use only the number of cores in one processor? That's the point of using

...

to run more threads in parallel than I have cores on my MacBook.

Depending on your budget, on AWS you can rent a machine from the miniscule t2.nano all the way up to the 128-core x1.32xlarge, with various

...

on memory vs CPU vs storage vs ... in between. All can run Haskell

On Mon, Jul 17, 2017 at 3:57 PM, David Turner wrote: them, tradeoffs programs.

...

There's also various options in Haskell for distributing your computation across multiple machines, if that sounds useful.

A good place to start is the pricing page which gives you a good overview of the options.

https://aws.amazon.com/ec2/pricing/on-demand/

Other cloud providers are available, of course.

Cheers,

D

_______________________________________________ Haskell-Cafe mailing list To (un)subscribe, modify options or view archives go to: http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe Only members subscribed via the mailman list are allowed to post.

_______________________________________________ Haskell-Cafe mailing list To (un)subscribe, modify options or view archives go to: http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe Only members subscribed via the mailman list are allowed to post.

_______________________________________________ Haskell-Cafe mailing list To (un)subscribe, modify options or view archives go to: http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe Only members subscribed via the mailman list are allowed to post.

Makes sense.​ Thanks D

Some links I like to to add to the thread: AWS lambda https://github.com/abailly/aws-lambda-haskell Google Cloud functions might be an option using GHCJS Heroku has been around for a long time: https://www.christopherbiscardi.com/2014/2/2/deploy-haskells-snap-on-heroku/ Greetings, Bram On Wed, Jul 19, 2017 at 5:25 AM, Mark Fine wrote:

The performance might not be what you're looking for (there are some more performant configurations but they might not be cost effective), but it's pretty easy to get a stack project running on Heroku - this script took ~3 minutes (dominated by installing GHC on the first push) to get a simple worker running in the cloud:

https://gist.github.com/mfine/b3dd21cfd03d1775084780f955cdc784#gistcomment-2...

Sample run:

~ $ stack new worker https://gist.githubusercontent.com/mfine/b3dd21cfd03d1775084780f955cdc784/ra... Downloading template "https://gist.githubusercontent.com/mfine/b3dd21cfd03d1775084780f955cdc784/ra..." to create project "worker" in worker/ ... Looking for .cabal or package.yaml files to use to init the project. Using cabal packages: - worker/worker.cabal

Selecting the best among 10 snapshots...

* Matches lts-8.23

Selected resolver: lts-8.23 Initialising configuration using resolver: lts-8.23 Total number of user packages considered: 1 Writing configuration to file: worker/stack.yaml All done. ~ $ ~ $ cd worker ~/worker $ ~/worker $ git init . Initialized empty Git repository in ~/worker/.git/ ~/worker $ ~/worker $ git add . ~/worker $ ~/worker $ git commit -m "initial" [master (root-commit) a19a152] initial 4 files changed, 84 insertions(+) create mode 100644 Procfile create mode 100644 main.hs create mode 100644 stack.yaml create mode 100644 worker.cabal ~/worker $ ~/worker $ heroku create --buildpack https://github.com/mfine/heroku-buildpack-stack.git Creating app... done, ⬢ serene-earth-29336 Setting buildpack to https://github.com/mfine/heroku-buildpack-stack.git... done https://serene-earth-29336.herokuapp.com/ | https://git.heroku.com/serene-earth-29336.git ~/worker $ ~/worker $ git push heroku master Counting objects: 6, done. Delta compression using up to 4 threads. Compressing objects: 100% (5/5), done. Writing objects: 100% (6/6), 1.61 KiB | 0 bytes/s, done. Total 6 (delta 0), reused 0 (delta 0) remote: Compressing source files... done. remote: Building source: remote: remote: -----> Haskell app detected remote: -----> 2017-07-19_02:58:20 Starting... remote: -----> 2017-07-19_02:58:20 Exporting config vars remote: -----> 2017-07-19_02:58:20 Installing libgmp-6.0.0a into cache remote: ######################################################################## 100.0% remote: -----> 2017-07-19_02:58:20 Restoring 6.0.0a files from cache remote: -----> 2017-07-19_02:58:20 Downloading stack-1.4.0 remote: ######################################################################## 100.0% remote: -----> 2017-07-19_02:58:21 Restoring stack-1.4.0 remote: -----> 2017-07-19_02:58:21 Substituting stack vars remote: -----> 2017-07-19_02:58:21 Running stack remote: Downloading lts-8.23 build plan ... remote: Downloaded lts-8.23 build plan. remote: Updating package index Hackage (mirrored at https://s3.amazonaws.com/hackage.fpcomplete.com/) ... remote: Selected mirror https://s3.amazonaws.com/hackage.fpcomplete.com/ remote: Downloading root remote: Selected mirror https://s3.amazonaws.com/hackage.fpcomplete.com/ remote: Downloading timestamp remote: Downloading snapshot remote: Downloading mirrors remote: Cannot update index (no local copy) remote: Downloading index remote: Updated package list downloaded remote: Populating index cache ... remote: Populated index cache. remote: Preparing to install GHC to an isolated location. remote: This will not interfere with any system-level installation. remote: Preparing to download ghc-8.0.2 ... remote: ghc-8.0.2: download has begun remote: ghc-8.0.2: 27.84 MiB / 107.55 MiB ( 25.89%) downloaded... remote: ghc-8.0.2: 54.10 MiB / 107.55 MiB ( 50.31%) downloaded... remote: ghc-8.0.2: 77.99 MiB / 107.55 MiB ( 72.52%) downloaded... remote: ghc-8.0.2: 101.87 MiB / 107.55 MiB ( 94.72%) downloaded... remote: ghc-8.0.2: 107.55 MiB / 107.55 MiB (100.00%) downloaded... remote: Downloaded ghc-8.0.2. remote: Unpacking GHC into /app/tmp/cache/.stack/programs/x86_64-linux/ghc-8.0.2.temp/ ... remote: Configuring GHC ... remote: Installing GHC ... remote: Installed GHC. remote: stack will use a sandboxed GHC it installed remote: For more information on paths, see 'stack path' and 'stack exec env' remote: To use this GHC and packages outside of a project, consider using: remote: stack ghc, stack ghci, stack runghc, or stack exec remote: [1 of 2] Compiling Main ( /app/tmp/cache/.stack/setup-exe-src/setup-mPHDZzAJ.hs, /app/tmp/cache/.stack/setup-exe-src/setup-mPHDZzAJ.o ) remote: [2 of 2] Compiling StackSetupShim ( /app/tmp/cache/.stack/setup-exe-src/setup-shim-mPHDZzAJ.hs, /app/tmp/cache/.stack/setup-exe-src/setup-shim-mPHDZzAJ.o ) remote: Linking /app/tmp/cache/.stack/setup-exe-cache/x86_64-linux/tmp-Cabal-simple_mPHDZzAJ_1.24.2.0_ghc-8.0.2 ... remote: worker-0.1.0.0: configure (exe) remote: Configuring worker-0.1.0.0... remote: worker-0.1.0.0: build (exe) remote: Preprocessing executable 'worker' for worker-0.1.0.0... remote: [1 of 1] Compiling Main ( main.hs, .stack-work/dist/x86_64-linux/Cabal-1.24.2.0/build/worker/worker-tmp/Main.o ) remote: Linking .stack-work/dist/x86_64-linux/Cabal-1.24.2.0/build/worker/worker ... remote: worker-0.1.0.0: copy/register remote: Installing executable(s) in remote: /app/.stack-work/install/x86_64-linux/lts-8.23/8.0.2/bin remote: Copying from /app/.stack-work/install/x86_64-linux/lts-8.23/8.0.2/bin/worker to /app/.local/bin/worker remote: remote: Copied executables to /app/.local/bin: remote: - worker remote: -----> 2017-07-19_03:01:26 Making stack binaries available remote: -----> 2017-07-19_03:01:26 Caching .stack-work remote: -----> 2017-07-19_03:01:26 Finished! remote: -----> Discovering process types remote: Procfile declares types -> worker remote: remote: -----> Compressing... remote: Done: 1.5M remote: -----> Launching... remote: Released v3 remote: https://serene-earth-29336.herokuapp.com/ deployed to Heroku remote: remote: Verifying deploy... done. To https://git.heroku.com/serene-earth-29336.git * [new branch] master -> master ~/worker $ ~/worker $ heroku ps:scale worker+1 Scaling dynos... done, now running worker at 1:Free ~/worker $ ~/worker $ heroku logs -t 2017-07-19T03:07:19.755903+00:00 heroku[worker.1]: Starting process with command `worker` 2017-07-19T03:07:20.553903+00:00 heroku[worker.1]: State changed from starting to up 2017-07-19T03:07:22.393957+00:00 app[worker.1]: Hello! 2017-07-19T03:07:23.397139+00:00 app[worker.1]: Hello! 2017-07-19T03:07:24.398803+00:00 app[worker.1]: Hello! 2017-07-19T03:07:25.400211+00:00 app[worker.1]: Hello! 2017-07-19T03:07:26.405199+00:00 app[worker.1]: Hello!

Mark

On Sun, Jul 16, 2017 at 1:20 PM, Dennis Raddle wrote:

...

I'm thinking of using cloud computing services for running my CPU-intensive parallel Haskell code, in order to get results faster. Instead of buying a faster computer, I would take advantage of existing services.

Note that my budget is small, so I need something that doesn't have a large minimum charge. I'm thinking $10 - 30 per month.

Are there any cloud services like that that run Haskell? D

_______________________________________________ Haskell-Cafe mailing list To (un)subscribe, modify options or view archives go to: http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe Only members subscribed via the mailman list are allowed to post.

_______________________________________________ Haskell-Cafe mailing list To (un)subscribe, modify options or view archives go to: http://mail.haskell.org/cgi-bin/mailman/listinfo/haskell-cafe Only members subscribed via the mailman list are allowed to post.