Building systems with etcd @ SCALE x14

SCALE x14

Transcript

1. Fault Tolerant Infrastructure Building Systems with etcd @coreoslinux @brandonphilips
2. Brandon Philips CTO, CoreOS github.com/philips
3. What is CoreOS?
4. What is CoreOS?
7. What is CoreOS?
9. The smartest way to run your container infrastructure. tectonic.com @tectonic
10. QUAY Secure hosting for private Docker repositories quay.io @quayio
11. Why build CoreOS?
12. you
13. you as a sw engineer
14. your with Ada.Text_IO; procedure Hello_World is use Ada.Text_IO; begin Put_Line("Hello, world!"); end; #include <stdio.h> int main() { printf("Hello, world!\n"); } package main import "fmt" func main() { fmt.Println("Hello, world!") }
15. your container image
16. your /bin/java /opt/app.jar /lib/libc
17. your /bin/python /opt/app.py /lib/libc
18. your com.example.app d474e8c57737625c
19. your d474e8c57737625c Signed By: Alice
20. you as an ops engineer
21. your
22. your com.example.webapp x3
23. your com.example.webapp x3
24. your ??? com.example.webapp x3
25. How do we do it?
26. architecture in practice cluster operations
27. worker kubelet worker kubelet worker kubelet scheduler & API worker kubelet w ku t worker kubelet
28. machine configuration OS operations
29. distributed configuration cluster operations
30. github.com/philips/hacks/tree/master/etcd- demos
31. etcd
32. /etc distributed
33. open source software failure tolerant durable watchable exposed via HTTP runtime reconfigurable
34. Data Store API -X GET Get Wait -X PUT Put Create CAS -X DELETE Delete CAD
35. clusters etcd basics
36. Typical Cluster Leader Follower
37. API etcd basics
38. fault tolerance etcd basics
39. Available Leader Follower
40. Available Leader Follower
41. Available Leader Follower
42. Unavailable Leader Follower
43. leader fault tolerance etcd basics
44. Available Leader Follower
45. Available Leader Follower
46. Temporarily Unavailable Leader Follower
47. Available Leader Follower
48. Unavailable Leader Follower
49. wal, snapshots, backups etcd durability
50. discovery, static etcd bootstrap
51. $ curl discovery.etcd.io/new?size=5 discovery.etcd.io/6eadeac2
52. discovery
53. discovery
54. discovery
55. Leader Follower discovery
56. live addition and removal etcd reconfig
57. Leader Follower
58. Leader Follower
59. Leader Follower
60. etcd apps
61. reboot locksmith etcd apps
64. Cluster Wide Reboot Lock • Need to reboot? Decrement the semaphore key atomically with etcd. • manager.Reboot() and wait... • After reboot increment the semaphore key in etcd atomically.
65. skydns etcd apps
66. vulcand etcd apps
68. confd etcd apps
69. pulling it together kubernetes
70. k8s/mesos/etc scheduler scheduling
71. getting work to servers scheduling
72. $ scp app host:/opt $ ssh host systemd-run /opt/app
73. $ scp app host:/opt $ ssh host systemd-run /opt/app
74. $ fab deploy:app
75. $ fab deploy:app
76. $ fab deploy:app
77. $ fab deploy:collector-app
78. $ fab deploy:collector-app
79. $ fab deploy:collector-app
80. $ fab deploy deploy:collector-app
81. $ fab lowest-loadaverage
82. $ fab lowest-loadaverage host1
83. $ fab lowest-loadaverage host1 $ fab -H host1 deploy:job
84. You Scheduler API Scheduler Machine(s)
85. while true { todo = diff(desState, curState) schedule(todo) }
86. while true { todo = diff(desState, curState) schedule(todo) }
87. while true { todo = diff(desState, curState) schedule(todo) }
88. while true { todo = diff(desState, curState) schedule(todo) }
89. dns, LBs, k8s labels services
90. flexible service discovery k8s labels
91. pod env=dev app=web pod env=test app=web pod env=prod app=web
92. pod env=dev app=web pod env=test app=web pod env=prod app=web service test.example.com select(env=dev,app=web) service beta.example.com select(env=test,app=web) OR select(env=prod,app=web) service example.com select(env=prod,app=web)
93. github.com/coreos/coreos-kubernetes
94. etcd.ngrok.io
95. worker kubelet worker kubelet scheduler & API
96. worker & API works on 1 node too
97. coreos.com/careers work with us
98. @coreoslinux @tectonicstack @brandonphilips thank you