Silo-Based Architectures For High Availability Applications

Slide 1

Slide 1 text

FOR HIGH AVAILABILITY APPLICATIONS SILO-BASED ARCHITECTURES Georgiana Gligor / Tekkie Consulting / @gbtekkie

Slide 2

Slide 2 text

@gbtekkie DevOpsDays Zürich 2 ✤ Geek. Mother. Do-er. ✤ on LAMP/LEMP stack since 2003 ✤ Architecture / DevOps consultant ✤ RomaniaPHP Organizer ✤ PhD Student @gbtekkie

Slide 3

Slide 3 text

@gbtekkie DevOpsDays Zürich 3 advantages and disadvantages silos: a possible approach the need for high availability what is high availability (HA)? AGENDA

Slide 4

Slide 4 text

No content

Slide 5

Slide 5 text

@gbtekkie DevOpsDays Zürich 5 https://youtu.be/MQm5BnhTBEQ

Slide 6

Slide 6 text

6 Software industry is built around anticipating change.

Slide 7

Slide 7 text

7 anticipate accommodate vs

Slide 8

Slide 8 text

TYPICAL APPLICATION

Slide 9

Slide 9 text

@gbtekkie DevOpsDays Zürich 9

Slide 10

Slide 10 text

No content

Slide 11

Slide 11 text

@gbtekkie DevOpsDays Zürich master Frontend Business Logic Frontend Frontend Browser internet Load balancer slave reads writes 11 ADJUSTING

Slide 12

Slide 12 text

@gbtekkie DevOpsDays Zürich master Frontend Business Logic Frontend Frontend Browser internet Load balancer slave reads writes 12 ADJUSTING redundancy

Slide 13

Slide 13 text

@gbtekkie DevOpsDays Zürich master Frontend Business Logic Frontend Frontend Browser internet Load balancer slave reads writes 13 ADJUSTING resilience

Slide 14

Slide 14 text

@gbtekkie DevOpsDays Zürich 14 TYPICAL LAYERING

Slide 15

Slide 15 text

@gbtekkie DevOpsDays Zürich 15 APPLICATION ARCHITECTURE

Slide 16

Slide 16 text

HIGH AVAILABILITY

Slide 17

Slide 17 text

@gbtekkie DevOpsDays Zürich 17 Ability to access the system: ✤ retrieve information ✤ alter information ✤ send new data AVAILABILITY

Slide 18

Slide 18 text

https:/ /flic.kr/p/dkasBz

Slide 19

Slide 19 text

@gbtekkie DevOpsDays Zürich 19 THE 9s DANCE Uptime Downtime (per year) 90.000 % 36.50 days one nine 99.000 % 3.65 days two nines 99.900 % 8.76 hrs three nines 99.950 % 4 hrs 23 mins 99.990 % 52.56 mins four nines 99.999 % 5.26 mins five nines

Slide 20

Slide 20 text

@gbtekkie DevOpsDays Zürich 20 THE 9s DANCE Uptime Downtime (per year) 90.000 % 36.50 days 99.000 % 3.65 days 99.900 % 8.76 hrs 99.950 % 4 hrs 23 mins Amazon SLA 99.990 % 52.56 mins four nines 99.999 % 5.26 mins five nines

Slide 21

Slide 21 text

@gbtekkie DevOpsDays Zürich 21 IMPACT $ 144,000 / hour 3600 $ 40 / sec * =

Slide 22

Slide 22 text

@gbtekkie DevOpsDays Zürich 22 USER BEHAVIOUR amazon facebook youtube Alexa Rank 6 3 2 daily time on site 12:07 mins 19:27 mins 23:44 mins daily pageviews / visitor 11.83 9.38 12.84 bounce rate 21 % 29 % 33 %

Slide 23

Slide 23 text

@gbtekkie DevOpsDays Zürich 23 HIGH AVAILABILITY TRIANGLE cost complexity risk

Slide 24

Slide 24 text

@gbtekkie DevOpsDays Zürich 24 DOWNTIME scheduled ‣ you unscheduled ‣ you ‣ others

Slide 25

Slide 25 text

@gbtekkie DevOpsDays Zürich 25 HAPPENS TO THE BEST

Slide 26

Slide 26 text

@gbtekkie DevOpsDays Zürich 26 MICHAEL JACKSON

Slide 27

Slide 27 text

H.A. SYSTEM CHARACTERISTICS

Slide 28

Slide 28 text

@gbtekkie DevOpsDays Zürich https://flic.kr/p/quMmFw NO SINGLE POINT OF FAILURE

Slide 29

Slide 29 text

@gbtekkie DevOpsDays Zürich https://flic.kr/p/RLKw8z RELIABLE CROSSOVER

Slide 30

Slide 30 text

@gbtekkie DevOpsDays Zürich DETECT FAILURES AS THEY OCCUR

Slide 31

Slide 31 text

@gbtekkie DevOpsDays Zürich 31 HA BEST PRACTICES 1. no single points of failure 2. stateless application design 3. automate infrastructure for consistency & reliability 4. clever monitoring and alerting 5. geographically distribute your machines 6. keep spare capacity to meet increasing demand

Slide 32

Slide 32 text

32 A man’s got to know his limitations. - Dirty Harry

Slide 33

Slide 33 text

SILOS

Slide 34

Slide 34 text

@gbtekkie DevOpsDays Zürich 34 TRY UPGRADE TO PHP7

Slide 35

Slide 35 text

@gbtekkie DevOpsDays Zürich 35 WHAT IS A SILO? ✤ frontend (SPAs, PWAs, etc) ✤ backend (e.g. PHP services) ✤ data (including cache) 1 silo = full setup of servers that deliver the end-to-end functionality

Slide 36

Slide 36 text

@gbtekkie DevOpsDays Zürich 36 WHAT IS A SILO?

Slide 37

Slide 37 text

@gbtekkie DevOpsDays Zürich 37 SILO-BASED ARCHITECTURE

Slide 38

Slide 38 text

@gbtekkie DevOpsDays Zürich 38 MULTIPLE CACHES

Slide 39

Slide 39 text

@gbtekkie DevOpsDays Zürich 39 A/B TESTING

Slide 40

Slide 40 text

@gbtekkie DevOpsDays Zürich 40 GEOGRAPHICAL DISTRIBUTION

Slide 41

Slide 41 text

@gbtekkie DevOpsDays Zürich 41 LIVE UPGRADES

Slide 42

Slide 42 text

@gbtekkie DevOpsDays Zürich 42 ADVANTAGES ✤ reuse familiar technology ✤ real A/B testing ✤ no BHUF requirements ✤ no disruption => brand loyalty ✤ lower Total Cost of Ownership ✤ simplify scalability

Slide 43

Slide 43 text

@gbtekkie DevOpsDays Zürich 43 DISADVANTAGES ✤ needs razor-sharp DevOps team ✤ small increase in hardware costs on kick-off ✤ adds complexity to the monitoring layer ✤ reconsider traceability ✤ different bug reproducing and hunting

Slide 44

Slide 44 text

@gbtekkie DevOpsDays Zürich 44 TAKEAWAYS

Slide 45

Slide 45 text

@gbtekkie DevOpsDays Zürich 45 ✤ build situational awareness with clever monitoring ✤ automate outage detection ✤ powerful A/B testing TAKEAWAYS

Slide 46

Slide 46 text

@gbtekkie DevOpsDays Zürich 46 FURTHER READING ✤ Wikipedia HA page ✤ OpenStack’s HA concepts ✤ Merge Hemo report from FDA ✤ USA Presidential Policy Directive 21 ✤ “Beyond Legacy Code” book ✤ TechCrunch’s summary of sites affected by Michael Jackson’s death ✤ Netflix lessons learned after AWS outage ✤ Netflix Chaos Monkey source code ✤ Brian Adler’s talk on “Architecting for High Availability and Multi-Cloud”

Slide 47

Slide 47 text

‹#› Questions? } Efficient architecture. Performance oriented. Enhanced with AI. [email protected]