Deployment in a new Cloud Era

Presentation about shipping and tooling

Transcript

1. lostpropertyhq.com Deployment in a new Cloud Era presentation about shipping and tooling / Created byRach Belaid @rachbelaid
2. Who I am. Python charmer Git Lover PostgreSQL Fan I build web products and love learning something new.
3. Version controle matters!! It keeps the history of the application.
4. Boom! magic rollback! What about servers? How do I version them? Unfortunately I cannot do: g i t r e s e t - - h a r d < p r e v i o u s >
5. What is an application? Is it just code? Or, does it include all its components?
6. This is a talk about shipping release. Warning: There is no silver bullet for all cases.
7. What's a Good release? Self Contained Easy to rollback Fast to deploy Easy to upgrade (less legacy) Easy to duplicate Testable in parallel
8. What's the cloud? Buzzword Flexible resources More resources on demande Pay for what you use
9. What to ship then? Source Code Python package Distribution Package LXC Images
10. Tools Python Wheels FPM Docker Packer Use tools or contribute but don't reinvent the wheel ;)
11. Python Wheels PEP 427 - A wheel is a ZIP-format archive.
12. Example: Create Wheels p i p i n s t a l l w h e e l p i p w h e e l - - w h e e l - d i r = / l o c a l / w h e e l s - r r e q u i r e m e n t s . t x t p i p w h e e l - - w h e e l - d i r = / l o c a l / w h e e l s . / a p p
13. Example: Install Wheels p i p i n s t a l l - - u s e - w h e e l - - n o - i n d e x - - f i n d - l i n k s = h t t p : / / m y f a n c y u r l /
14. Downsides Not self-contained. Just code. Not easy to rollback. Portability.
15. Distro Package via FPM
16. Fundamental Principle: “ IF FPM IS NOT HELPING YOU MAKE PACKAGES EASILY, THEN THERE IS A BUG IN FPM.”
17. Example: Create Package f p m - s p y t h o n - t d e b . / a p p / s e t u p . p y f p m - s d i r - t d e b a p p _ f o l d e r _ w i t h _ v e n v /
18. Upsides Easy to rollback. Make migration easier. Wrap your dependencies.
19. Downsides Not self-contained (enough). Just code. Platform dependant.
20. What about the big picture
21. Application are more complex and not just code they have dependencies!! Redis Postgres System Libs
22. Configuration managment?
23. I love sensible configuration managment Salt Ansible
24. But they are not easy to rollback ... even if you keep it in a VCS. It's not enough.
25. Immutable servers From Credit: Martin Fowler's blog post on Immutable Server
26. Immutable servers To Credit: Martin Fowler's blog post on Immutable Server
27. Versionning Server State!
28. LXC, made easy with Docker
29. What's docker Docker, the Linux Container Runtime, an application to pack, ship and run any application as a lightweight container
30. Why container matter? Content Agnostic Hardware Agnostic Content isolation Automation Separation of duties
31. Get Started g i t c l o n e h t t p s : / / g i t h u b . c o m / d o t c l o u d / d o c k e r . g i t c d d o c k e r v a g r a n t u p v a g r a n t s s h s u d o d o c k e r
32. Docker Awesomeness Versionning Push / Pull Remote/Private Registery Logs Diff REST API Support a docker inside a docker
33. Downsides Kernel version Auto-scalling
34. Images, made easy via Packer packer.io
35. What's packer Packer is a tool for creating identical machine images for multiple platforms from a single source configuration.
36. Simple JSON to describe 3 steps: Provisioning (shell, files, Salt, Chef) Build (AWS, DigitalOcean, VMWare, OpenStack ) Post-Process (Vagrant, Compress)
37. Examples { " b u i l d e r s " : [ { " t y p e " : " a m a z o n - e b s " , " r e g i o n " : " u s - e a s t - 1 " , " s o u r c e _ a m i " : " a m i - e 5 0 e 8 8 8 c " , " i n s t a n c e _ t y p e " : " t 1 . m i c r o " , " s s h _ u s e r n a m e " : " u b u n t u " , " a m i _ n a m e " : " p a c k e r - e x a m p l e { { t i m e s t a m p } } " } ] , " p r o v i s i o n e r s " : [ { " t y p e " : " s h e l l " , " i n l i n e " : [ " e c h o ' I c a n c h a i n P r o v i s i o n e r s ' " ] } , { " t y p e " : " s a l t - m a s t e r l e s s " , " b o o t s t r a p _ a r g s " : " g i t v 0 . 1 6 . 0 " , " l o c a l _ s t a t e _ t r e e " : " p a t h / t o / s t a t e s " , " l o c a l _ p i l l a r _ r o o t s " : " p a t h / t o / p i l l a r " } ] }
38. What to use where There is no rules. They all can play well together (with a shoehorn)
39. private libs, dependencies, pur python apps Wheels
40. Migration, datacenter, no VM FPM
41. Recent Kernel (or patched), datacenter, Cloud Provider Docker
42. Cloud provider, Private Cloud, Virtualized Datacenter: AWS, Digital Ocean. Packer
43. A Final Word. Have Fun and Pick your tools.