Don’t think about hardware / OS Don’t think (much) about network DNS handled for you SSL/TLS termination Auto-scaling (depending on provider/platform) dev/stage/prod environments are easy
Foundry (Pivotal) ‡ FP Haskell Centre (FP Complete) ‡ . . . various others. † free tier available ‡ runs on free software i.e. you can run an instance yourself
for major frameworks (Snap, Yesod, scotty, HappStack) Manifest URLs: http://www.haskell.org/haskellwiki/Web/Cloud Currently a bit broken (I have submitted a PR to fix)
for major frameworks (Snap, Yesod, scotty, HappStack) Manifest URLs: http://www.haskell.org/haskellwiki/Web/Cloud Currently a bit broken (I have submitted a PR to fix) Manage using rhc command-line tool.
I know of). So it might be nice. . . ? There is no free tier. #sadface Haven’t tried it. . . but I would love to hear about your experiences if you have/do!
env vars. Differences between providers. More services (databases, metrics, etc) = more config. I’m chipping away on a small library to abstract (some of) this.
env vars. Differences between providers. More services (databases, metrics, etc) = more config. I’m chipping away on a small library to abstract (some of) this. Deployment:
env vars. Differences between providers. More services (databases, metrics, etc) = more config. I’m chipping away on a small library to abstract (some of) this. Deployment: Some PaaS run your app according to some convention (e.g. OpenShift)
env vars. Differences between providers. More services (databases, metrics, etc) = more config. I’m chipping away on a small library to abstract (some of) this. Deployment: Some PaaS run your app according to some convention (e.g. OpenShift) Some PaaS require you to tell it how to run your app (e.g. Heroku, via Procfile)
frameworks take up several hundred MB. I have done some work on OpenShift cartridges to improve the situation. I have some more ideas. Pay $$$ for more disk.
frameworks take up several hundred MB. I have done some work on OpenShift cartridges to improve the situation. I have some more ideas. Pay $$$ for more disk. Low memory
frameworks take up several hundred MB. I have done some work on OpenShift cartridges to improve the situation. I have some more ideas. Pay $$$ for more disk. Low memory e.g. can’t build haskell-src-exts on small OpenShift gear
frameworks take up several hundred MB. I have done some work on OpenShift cartridges to improve the situation. I have some more ideas. Pay $$$ for more disk. Low memory e.g. can’t build haskell-src-exts on small OpenShift gear Solution: ship more cartridge variants with more libs?
frameworks take up several hundred MB. I have done some work on OpenShift cartridges to improve the situation. I have some more ideas. Pay $$$ for more disk. Low memory e.g. can’t build haskell-src-exts on small OpenShift gear Solution: ship more cartridge variants with more libs? Pay $$$ for more memory.
sets. OpenShift cartridges might support Stackage soon. Docker Easier to set up container images. Layered images = shared base images, smaller framework/app images.
sets. OpenShift cartridges might support Stackage soon. Docker Easier to set up container images. Layered images = shared base images, smaller framework/app images. Lots of Docker hosting providers already.
sets. OpenShift cartridges might support Stackage soon. Docker Easier to set up container images. Layered images = shared base images, smaller framework/app images. Lots of Docker hosting providers already. OpenShift v3 is embracing Docker.
sets. OpenShift cartridges might support Stackage soon. Docker Easier to set up container images. Layered images = shared base images, smaller framework/app images. Lots of Docker hosting providers already. OpenShift v3 is embracing Docker. First-class support for Haskell in popular PaaS providers?
sets. OpenShift cartridges might support Stackage soon. Docker Easier to set up container images. Layered images = shared base images, smaller framework/app images. Lots of Docker hosting providers already. OpenShift v3 is embracing Docker. First-class support for Haskell in popular PaaS providers? Hopefully! FP Complete seem to be on their own, for now.
on OpenShift blog post by Kate Miller: http://is.gd/7eRZDY OpenShift cartridge development: https://github.com/accursoft/Haskell-Cloud Heroku: https://www.heroku.com/ Heroku buildpack development: https://github.com/begriffs/heroku-buildpack-ghc Keter: https://github.com/snoyberg/keter
licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Slides https://github.com/frasertweedale/talks Email [email protected] Twitter @hackuador