DevOps

DevOps introduction

Transcript

1. DevOps Introductio Ops' job is NOT to keep the site stable and fast Ops' job is to enable the business (this is dev's job too)
2. Who am i ?  Systems Administrator @  DevOps  Infrastructure as a code ”Puppet”  Cloud Computing  …  @EslamElHusseiny
3. Dev or Ops ? 6 days after a successful deployment, server experiences troubles ”high load, resources leak”
4. Dev or Ops ? There is an issue with build server. Code isn't building.
5. Dev or Ops ? Software is deployed to the test environment. You can't login to application.
6. Dev vs. Ops
7. Dev says ...
8. Ops replies ... ”It's not my server, it's your code”
9. Role of Devs and Ops
10. Role of Devs and Ops  Dev  Create changes  Add / modify features  Ops  Maintain stability  Create / enhance services
11. Role of Devs and Ops  Dev  Create changes  Add / modify features  Ops  Maintain stability  Create / enhance services
12. So ?  The problem  Disconnect between groups  Devs don't deploy consistent software  Ops motivated to resist changes  Devolpment process is Agile  Operations process is Static
13. Our Goal We are all here to ENABLE business
14. change  Is required for business  Doesn't mean outage  Should be exploited and not feared  Shouldn't mean fire fighting
15. Rescue DevOps To The Rescue !
16. What is DevOps then ?  ”DevOps is an approach to bridge the gap between agile software development and operations” ­ agileweboperations.com  Collaborative mindset of Devs and Ops  C.A.M.S
17. Culture
18. Culture  Relationships  Engage early, engage often  Be open  Eradicate ”last­mile syndrome”
19. Culture  Communication  Talk is cheap (get out of your chair)  Involve each other in core proccesses and decisions  Ask Questions  Don't say ”No”
20. Automation
21. Automation  What ca be automated ?  Builds  Deployments  Testing  Monitoring  Self­healing  System configuration
22. Metrics
23. Metrics  Capture, learn, improve.  Assist in:  Capacity planning  Trend analysis  Fault finding
24. Metrics
25. Sharing
26. Sharing  Share ideas  Share metrics  Ops : Give devs shell access  Devs : see what technology ca be leveraged
27. DevOps life cycle
28. Before  Talk about functional requirements  Talk about non­functional requirements  Security  Backups  Availability  Upgradeability  Configuration management  Monitoring / Metrics  Logging
29. During  Communication  Source control  Automate builds  Automate tests  Automate deployments (Dev, Test, and Prod)  Collate App and System metrics
30. After  Release  Continue to run tests  Monitor application and system  Issues (might happen)  Post meetings
31. What are we doing ?
32. Puppet
33. Puppet  Configuration management tool  Open Source  Infrastructure as a code
34. Puppet
35. Puppet
36. node default { include ssh include sudo include ntp } node product.dev inherits default { include nginx include tomcat5 } node product.test inherits default{ include nginx include tomcat6 }
37. class sudo { package {'sudo': ensure => latest, } file {'/etc/sudoers': owner => root, group => root, mode => '0440', source => 'puppet:///files/sudo/sudoers', require => Package['sudo'], } }
38. Puppet
39. Let's get hands on It's deploymnt time
40. Monitoring
41. Monitoring
42. Problem ?  No one noticed  No one got alerted  No automatic recovery
43. Probelm ?  what is your application doing right now?  do you know when it fails?  failure means customers lose trust  failure means customers go elsewhere  failure means you lose money  application = providing value
44. Solution  Monitoring  Metrics  Logging
45. Monitoring means Is the application available?
46. Examples
47. Monitoring is this service currently providing value? is this service consuming too many resources?
48. Examples
49. Monit check process unicorn with pidfile /var/run/unicorn/unicorn.pid start program = "/etc/init.d/unicorn start" stop program = "/etc/init.d/unicorn stop" if mem is greater than 300.0 MB for 1 cycles then restart if cpu is greater than 50% for 2 cycles then alert if cpu is greater than 80% for 3 cycles then restart group unicorn
50. Metrics  measurements  historical data  graphs
51. Metrics how many customers are on my site? how many customers were on my site yesterday?
52. Metrics how slow is paypal's api? how slow was paypal's api yesterday?
53. Metrics how much memory is available on my servers? how much has memory usage grown over four weeks?
54. Metrics number of open database connections number of redis commands number of 500 errors rate of HTTP requests number of HTTP connections
55. Metrics median response time number of failed resque jobs number of twitter followers 99th percentile github api response time 95th percentile mysql query time
56. Examples
57. Munin
58. Graphite
59. Metrics as a service
60. Logging
61. Logstash log inputs process outputs
62. Graylog2
63. Graylog2
64. Summary own your monitoring own your metrics own your logging none of them is optional