Slide 1

Slide 1 text

production: an owner's manual

Slide 2

Slide 2 text

hello!

Slide 3

Slide 3 text

broken computers

Slide 4

Slide 4 text

No content

Slide 5

Slide 5 text

getting sidetracked now so sorry* * not sorry

Slide 6

Slide 6 text

No content

Slide 7

Slide 7 text

No content

Slide 8

Slide 8 text

No content

Slide 9

Slide 9 text

back to serious business

Slide 10

Slide 10 text

!

Slide 11

Slide 11 text

No content

Slide 12

Slide 12 text

a production system is a system that serves real users

Slide 13

Slide 13 text

the goal of operations is to ensure services are reliable

Slide 14

Slide 14 text

in order to provide a good user experience

Slide 15

Slide 15 text

No content

Slide 16

Slide 16 text

failure

Slide 17

Slide 17 text

app

Slide 18

Slide 18 text

app linux kernel cpu dram disk network power supply switches load balancer dns submarine cables routers fiber

Slide 19

Slide 19 text

app linux kernel the cloud

Slide 20

Slide 20 text

No content

Slide 21

Slide 21 text

• cosmic rays • disk failure • power outages • software bugs • ...

Slide 22

Slide 22 text

entropy

Slide 23

Slide 23 text

No content

Slide 24

Slide 24 text

capacity

Slide 25

Slide 25 text

No content

Slide 26

Slide 26 text

No content

Slide 27

Slide 27 text

No content

Slide 28

Slide 28 text

cascading failure

Slide 29

Slide 29 text

No content

Slide 30

Slide 30 text

system design

Slide 31

Slide 31 text

redundancy

Slide 32

Slide 32 text

"

Slide 33

Slide 33 text

scale

Slide 34

Slide 34 text

No content

Slide 35

Slide 35 text

"

Slide 36

Slide 36 text

p1 m3 c1 m2 m1 p2 c2

Slide 37

Slide 37 text

data storage

Slide 38

Slide 38 text

"

Slide 39

Slide 39 text

"

Slide 40

Slide 40 text

protocols

Slide 41

Slide 41 text

No content

Slide 42

Slide 42 text

monitoring

Slide 43

Slide 43 text

many components many req/s

Slide 44

Slide 44 text

No content

Slide 45

Slide 45 text

measure all the things?

Slide 46

Slide 46 text

✅ ⏱

Slide 47

Slide 47 text

golden signals • latency • traffic • errors • saturation

Slide 48

Slide 48 text

golden signals • latency • traffic • errors • saturation

Slide 49

Slide 49 text

golden signals • latency • traffic • errors • saturation

Slide 50

Slide 50 text

golden signals • latency • traffic • errors • saturation

Slide 51

Slide 51 text

golden signals • latency • traffic • errors • saturation 0 - 50 [1620]: ∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎∎ (74.55%) 50 - 100 [ 447]: ∎∎∎∎∎∎∎∎∎∎ (20.57%) 100 - 150 [ 49]: ∎ (2.25%) 150 - 200 [ 15]: (0.69%) 200 - 250 [ 15]: (0.69%) 250 - 300 [ 10]: (0.46%) 300 - 350 [ 6]: (0.28%) 350 - 400 [ 1]: (0.05%) 400 - 450 [ 0]: (0.00%) 450 - 500 [ 4]: (0.18%)

Slide 52

Slide 52 text

golden signals • latency • traffic • errors • saturation

Slide 53

Slide 53 text

saturation traffic latency errors

Slide 54

Slide 54 text

No content

Slide 55

Slide 55 text

humans

Slide 56

Slide 56 text

No content

Slide 57

Slide 57 text

oops, deleted the database

Slide 58

Slide 58 text

bad human!

Slide 59

Slide 59 text

why does this button even exist?

Slide 60

Slide 60 text

app linux kernel cpu dram disk network power supply switches load balancer dns submarine cables routers fiber

Slide 61

Slide 61 text

app linux kernel cpu dram disk network power supply switches load balancer dns submarine cables routers fiber humans

Slide 62

Slide 62 text

app linux kernel cpu dram disk network power supply switches load balancer dns submarine cables routers fiber humans h u m a n s

Slide 63

Slide 63 text

epic failure is almost always systemic

Slide 64

Slide 64 text

failure

Slide 65

Slide 65 text

recap

Slide 66

Slide 66 text

• a production system serves real users • users like things that work and are fast • epic failure is almost always systemic

Slide 67

Slide 67 text

thx @igorwhilefalse

Slide 68

Slide 68 text

No content