Who owns the Service Level?

by Takeshi Kondo

Slide 1

Slide 1 text

Who owns the Service Level? Takeshi Kondo / @chaspy 2022/05/15 SRE NEXT 2022

Slide 2

Slide 2 text

Who am I chaspy chaspy_ Engineering Manager, Site Reliability at Recruit Co., Ltd. Takeshi Kondo https://chaspy.me

Slide 3

Slide 3 text

લఏɿϓϩμΫτ঺հ - ελσΟα ϓϦ

Slide 4

Slide 4 text

Hello, again! SRE NEXT

Slide 5

Slide 5 text

ࠓ೔࿩͢͜ͱ SRE Λ࣮ݱ͢ΔͨΊʹඞཁͳ͜ͱ

Slide 6

Slide 6 text

Tl;dr / SRE Λ࣮ݱ͢ΔͨΊʹඞཁͳ͜ͱ • ࢦඪΛݩʹվળ͠ଓ͚ΔจԽ • ࣗݾ׬݁ԽΛࢧ͑Δ Platform • ༏ઌॱҐΛมߋ͢ΔͨΊͷٕज़ઓུ

Slide 7

Slide 7 text

Agenda • લఏɿSRE Λ࣮ݱ͢Δͱ͸Ͳ͏͍͏͜ͱ͔ • ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱࣄۀʹ͓͚Δ SRE • 2020೥ʮSLO ReviewʯͰͷֶͼͱࣦഊ • SRE ͱٕज़ઓུ • ·ͱΊͱࠓޙ

Slide 8

Slide 8 text

Agenda • લఏɿSRE Λ࣮ݱ͢Δͱ͸Ͳ͏͍͏͜ͱ͔ • ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱࣄۀʹ͓͚Δ SRE • 2020೥ʮSLO ReviewʯͰͷֶͼͱࣦഊ • SRE ͱٕज़ઓུ • ·ͱΊͱࠓޙ

Slide 9

Slide 9 text

લఏ෦෼͸3݄ͷ6ࣾ߹ಉษڧձͱಉ͡಺༰Ͱ͓͞Β͍͠·͢

Slide 10

Slide 10 text

SRE Λ࣮ݱ͢Δͱ͸

Slide 11

Slide 11 text

։ൃνʔϜ͕৴པੑΛ ίϯτϩʔϧ͢Δ Capability Λ ਎ʹ͚͍ͭͯΔ͜ͱ

Slide 12

Slide 12 text

ͦ΋ͦ΋ Site Reliability Engineering ͱ͸: Not like this • αʔϏε͕ʮߴ͍৴པੑ (ʹ100%)ʯΛอ͍ͬͯΔ͜ͱ • SLI/SLO ΛकΕ͍ͯΔ͜ͱ • ΦϯίʔϧϩʔςʔγϣϯΛ։ൃνʔϜͰߦ͏͜ͱ https://github.com/twitter/twemoji

Slide 13

Slide 13 text

ͦ΋ͦ΋ Site Reliability Engineering ͱ͸: Like this! • αʔϏε͕ʮϢʔβ͕ظ଴͢Δ৴པੑʯΛอ͍ͬͯΔ͜ͱ • SLI/SLO Λઃఆ͠ɺඇػೳཁ݅ͱػೳཁ݅ͷ༏ઌ౓ܾఆͷ ࢦඪͱͯ͠׆༻͍ͯ͠Δ • SLO ҧ൓͕ൃੜͨ͠ͱ͖ʹద੾ʹରॲͰ͖ΔΑ͏ͳϞχλ Ϧϯάํ๏ͱϙϦγʔ͕νʔϜͰಉҙ͞Ε͍ͯΔ • ্ه͕ఆظతʹݟ௚͞Ε͍ͯΔ https://github.com/twitter/twemoji

Slide 14

Slide 14 text

։ൃνʔϜ͕৴པੑΛίϯτϩʔϧ͢Δ Capability Λ਎ʹ͚ͭΔ: Like this! SRE ։ൃ νʔϜ ։ൃνʔϜͷ৴པੑʹ ؔ͢Δ Capability औಘ Λࢧԉ͢Δ ࣗ෼ͨͪͷαʔϏεͷ ৴པੑΛࣗ෼ͨͪͰί ϯτϩʔϧͰ͖͍ͯΔ

Slide 15

Slide 15 text

Team Topologies • 4ͭͷνʔϜύλʔϯ • Stream Aligned • Platform • Enabling • Complicated Subsystem • 3ͭͷίϛϡχέʔγϣϯύλʔϯ • Collaboration • X as a Service • Facilitation https://pub.jmam.co.jp/book/b593881.html

Slide 16

Slide 16 text

։ൃνʔϜ͕৴པੑΛίϯτϩʔϧ͢Δ Capability Λ਎ʹ͚ͭΔ: Like this! SRE ։ൃ νʔϜ ։ൃνʔϜͷࣗݾ׬݁ԽΛ ࢧ͑ΔϓϥοτϑΥʔϜͱ จԽΛ࡞Δ Platform Team Enabling Team Stream Aligned Team ࣗ෼ͨͪͰඞཁͳ΋ͷΛ ࣗ෼ͨͪͰ༻ҙͰ͖Δ = self-contained / ࣗݾ׬݁Խ

Slide 17

Slide 17 text

• ࢧԉ͸Ұ࣌తͰɺ௕ظԽͤͯ͞͸ͳΒͳ͍ ༨ஊ: Enabling ͱ͍͏୯ޠʹ͸ϙδςΟϒɾωΨςΟϒ྆໘͋Δ https://ja.wikipedia.org/wiki/%E3%82%A4%E3%83%8D%E3%83%BC%E3%83%96%E3%83%AA%E3%83%B3%E3%82%B0

Slide 18

Slide 18 text

SRE Team ͷ Vision / Mission / Values https://blog.studysapuri.jp/entry/sre-vision-mission-values

Slide 19

Slide 19 text

Mission ࣗݾ׬݁νʔϜ͕ϓϩμΫ τΛૉૣ҆͘શʹಧ͚ଓ͚ ΔͨΊͷϓϥοτϑΥʔϜ ͱจԽΛ࡞Δ

Slide 20

Slide 20 text

ͳͥࣗݾ׬݁Խ͕ॏཁ͔ SRE ։ൃ νʔϜ ։ൃνʔϜͷࣗݾ׬݁ԽΛ ࢧ͑ΔϓϥοτϑΥʔϜͱ จԽΛ࡞Δ Platform Team Enabling Team Stream Aligned Team ࣗ෼ͨͪͰඞཁͳ΋ͷΛ ࣗ෼ͨͪͰ༻ҙͰ͖Δ = self-contained / ࣗݾ׬݁Խ

Slide 21

Slide 21 text

ͳͥࣗݾ׬݁Խ͕ॏཁ͔: Not “VS”, but “And” • Dev vs and Ops • Ϣʔβ͔Βߴ଎ʹϑΟʔυόοΫΛಘΔ (DevOps) • Dev vs and Infrastructure • ηϧϑαʔϏεͰߏஙͯ͠ϦʔυλΠϜ୹ॖ • Productivity vs and Reliability • ੜ࢈ੑͱ৴པੑ͸૬ޓʹґଘ͢Δ

Slide 22

Slide 22 text

ͳͥࣗݾ׬݁Խ͕ॏཁ͔ • ࣗݾ׬݁νʔϜͳΒ • ຊ൪؀ڥ͔ΒϑΟʔυόοΫΛಘଓ͚ΒΕΔ • ࠷খݶͷίϛϡχέʔγϣϯͰߴ଎ͳҙࢥܾఆ͕Ͱ͖Δ • ػೳཁٻ͚ͩͰͳ͘ඇػೳཁٻʹ΋Ԡ͑Δ͜ͱ͕Ͱ͖Δ

Slide 23

Slide 23 text

• ։ൃνʔϜ͕৴པੑΛίϯτϩʔϧ͢Δ Capability Λ਎ʹ͚ͭͯ ͍Δ͜ͱ • ։ൃνʔϜ͕”ࣗݾ׬݁Խ”͍ͯ͠Δঢ়ଶ • SRE νʔϜ͸͜ΕΛϓϥοτϑΥʔϜͱจԽৢ੒Ͱࢧ͑Δ • ͜ΕΛ࣮ݱ͢Δʹ͸ϓϩμΫτ։ൃʹด͡ͳ͍ଟ༷ͳࢹ఺͕ඞཁ • Ϣʔβͷظ଴஋Λ஌Δ / Product Management • ߴ͍։ൃੜ࢈ੑ / Development Skills • ඇػೳཁٻʹͲΕ͚ͩίετΛ͔͚Δ͔ / Business Development ·ͱΊɿSRE Λ࣮ݱ͢Δͱ͸

Slide 24

Slide 24 text

Agenda • લఏɿSRE Λ࣮ݱ͢Δͱ͸Ͳ͏͍͏͜ͱ͔ • ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱࣄۀʹ͓͚Δ SRE • 2020೥ʮSLO ReviewʯͰͷֶͼͱࣦഊ • SRE ͱٕज़ઓུ • ·ͱΊͱࠓޙ

Slide 25

Slide 25 text

ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱSRE ~ @chaspy ೖࣾޙ • 2018: @chaspy ೖࣾ • 2019: Application Platform Λ Kubernetes ΁Ҡ؅ • 2020: Microservices Readiness ͷ੔උ • αʔϏεΦʔφʔγοϓͷࡦఆ • Design Doc / Production Readiness Checklist • Self-services Infrastructure (terraform monorepo) • SLI/SLO • 2021: SLI/SLO ӡ༻Λ։ൃνʔϜʹ׬શҠৡ Platform Team ͱͯ͠ Platform Λ࡞͍ͬͯΔ Enabling Team ͱͯ͠ ։ൃ૊৫ʹ SLI/SLO ͳͲͷΧϧνϟʔৢ੒

Slide 26

Slide 26 text

ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱSRE ~ 2021೥ • COVID-19 ྲྀߦɺΞΫηε਺૿େ • Platform ͷਐԽ • Terraform monorepo • Loadtest Platform • GitHub Actions ʹΑΔ monorepo CI ෼཭ • ૊৫ͷมԽ • ٕज़ઓུάϧʔϓൃ଍ • ࣄۀҠ؅ʹΑΓϦΫϧʔτ΁స੶ɺQuipper ೔ຊࢧళਫ਼ࢉ • chaspy EM ೚༻

Slide 27

Slide 27 text

ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱSRE ~ 2022೥ • ৽͍͠ SRE ࣮ફͷܗ • Partially Embedded SRE • Partially Enabling SRE in development team

Slide 28

Slide 28 text

૊৫ن໛ͷਪҠ ։ൃऀ 35 53 54 73 114 43& 4 5 7 7 7 ։ൃऀ͸ελσΟαϓϦɾQuipper ྆ํͷɺWeb Engineer (frontend&backend) ͷ਺ɻNative ͸আ֎͍ͯ͠Δɻ 2022೥͔Β͸ۀ຿ҕୗͷํ΋Χ΢ϯτ͍ͯ͠Δɻ2021೥Ҏલ΋ۀ຿ҕୗͷํͱ࢓ࣄ͸͍ͯͨ͠ɻ

Slide 29

Slide 29 text

ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱSRE • ͍ͣΕͷ࣌୅΋ Platform Team ͱ Enabling Team ͷৼΔ෣ ͍Λ͍ͯ͠Δ • ಛʹ2019೥͔Β͸ʮࣗݾ׬݁ԽʯΛςʔϚʹɺ͓ئ͍͞Ε Δ͜ͱΛۃྗݮΒͤΔ Platform Λ࡞͖ͬͯͨ • ಉ࣌ʹ։ൃνʔϜͷʮจԽΛͭ͘Δʯ͜ͱʹ౿ΈࠐΈɺSLI/ SLO Λݟ͍ͯ͘จԽΛ૊৫ʹৢ੒ͨ͠ • →ʮSLO Reviewʯat SRE NEXT 2020

Slide 30

Slide 30 text

Agenda • લఏɿSRE Λ࣮ݱ͢Δͱ͸Ͳ͏͍͏͜ͱ͔ • ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱࣄۀʹ͓͚Δ SRE • 2020೥ʮSLO ReviewʯͰͷֶͼͱࣦഊ • SRE ͱٕज़ઓུ • ·ͱΊͱࠓޙ

Slide 31

Slide 31 text

ʮSLO Reviewʯat SRE NEXT 2020 • ։ൃ૊৫ʹ SLO Λ Review ͍ͯ͘͠จԽΛ࡞ͬͨऔΓ૊Έ • 2 Product, 15 Team ʹϘτϜΞοϓͰಋೖ • ొΓํͷ4εςοϓ • γεςϜͱ૊৫ͷΦʔφʔγοϓΛܾΊΔ • 1ਓͰϓϩηεΛ·Θ͠ɺ࣮ݱํ๏Λཱ֬͢Δ • Developer ͱҰॹʹ SLI/SLO Λఆٛ͠ɺϓϩηεΛ·Θ͢ • Error Budget Policy Λఆٛͯ͠ߦಈ͢Δ(ະ࣮ݱ) • ಘֶͨͼ • ඪ४Խ͞Εͨ SLI Λఏڙ͢Δ • ઃఆ͸ૣ͍ஈ֊ͰίʔυԽ͢Δ • ֶशۂઢΛٸޯ഑ʹ͢Δ https://blog.studysapuri.jp/entry/2020/01/30/slo-review

Slide 32

Slide 32 text

Α͔ͬͨ఺ • ։ൃ૊৫ʹ SLO Λ Review ͍ͯ͘͠จԽΛ࡞ͬͨऔΓ૊Έ • 2 Product, 15 Team ʹϘτϜΞοϓͰಋೖ • ొΓํͷεςοϓ • γεςϜͱ૊৫ͷΦʔφʔγοϓΛܾΊΔ • 1ਓͰϓϩηεΛ·Θ͠ɺ࣮ݱํ๏Λཱ֬͢Δ • Developer ͱҰॹʹ SLI/SLO Λఆٛ͠ɺϓϩηεΛ·Θ͢ • Error Budget Policy Λఆٛͯ͠ߦಈ͢Δ(ະ࣮ݱ) • ಘֶͨͼ • ඪ४Խ͞Εͨ SLI Λఏڙ͢Δ • ઃఆ͸ૣ͍ஈ֊ͰίʔυԽ͢Δ • ֶशۂઢΛٸޯ഑ʹ͢Δ https://blog.studysapuri.jp/entry/2020/01/30/slo-review ։ൃνʔϜͷೝ஌ෛՙΛపఈత ʹԼ͛Δ͜ͱʹͩ͜Θͬͨ ໨తෆ࣮֬ੑͷ௿ݮͷͨΊ ϑΟʔυόοΫαΠΫϧΛճͨ͠

Slide 33

Slide 33 text

Α͘ͳ͔ͬͨ఺ʁ • ։ൃ૊৫ʹ SLO Λ Review ͍ͯ͘͠จԽΛ࡞ͬͨऔΓ૊Έ • 2 Product, 15 Team ʹϘτϜΞοϓͰಋೖ • ొΓํͷεςοϓ • γεςϜͱ૊৫ͷΦʔφʔγοϓΛܾΊΔ • 1ਓͰϓϩηεΛ·Θ͠ɺ࣮ݱํ๏Λཱ֬͢Δ • Developer ͱҰॹʹ SLI/SLO Λఆٛ͠ɺϓϩηεΛ·Θ͢ • Error Budget Policy Λఆٛͯ͠ߦಈ͢Δ(ະ࣮ݱ) • ಘֶͨͼ • ඪ४Խ͞Εͨ SLI Λఏڙ͢Δ • ઃఆ͸ૣ͍ஈ֊ͰίʔυԽ͢Δ • ֶशۂઢΛٸޯ഑ʹ͢Δ https://blog.studysapuri.jp/entry/2020/01/30/slo-review ͳͥ͏·͍͔͘ͳ ͔ͬͨͷ͔ʁ

Slide 34

Slide 34 text

ʢ࠶ܝʣSRE Λ࣮ݱ͢Δͱ͸Կ͔ • ։ൃνʔϜ͕৴པੑΛίϯτϩʔϧ͢Δ Capability Λ਎ʹͭ ͚͍ͯΔ͜ͱ • ʮࢦඪΛݩʹߦಈ͕Ͱ͖Δʯ͜ͱ https://blog.studysapuri.jp/entry/2020/01/30/slo-review 4-*4-0 &SSPS#VEHFU 1PMJDZ ։ൃνʔϜ ఆٛ͢Δ / ఆظతʹݟ௚͢ Policy ʹैͬͯߦಈ͢Δ

Slide 35

Slide 35 text

Slide 36

Slide 36 text

Slide 37

Slide 37 text

ͳͥ”ߦಈ͢Δ”·ͰࢸΒͳ͔ͬͨͷ͔ • ౰࣌ɺSLO ҧ൓࣌ͷΞΫγϣϯ͸ Product Manager / Team ʹҠৡ͍ͯͨ͠ • ·ͬͨ͘Կ΋Ͱ͖ͳ͔ͬͨΘ͚Ͱ͸ͳ͍ • ΋ͱ΋ͱνʔϜʹ༧ࢉͷ͋ΔɺվળͷͨΊͷ࣌ؒͰͰ͖Δ͜ͱ͔͠Ͱ͖ͳ͔ͬͨʢִि1೔ʣ • QB Day ͱݺ͹ΕΔ • Τϥʔʹର͢Δ௚઀తͳରॲɺܰඍͳ Performance վળͳͲ • ΞʔΩςΫνϟมߋɺΠϯϑϥ෼཭ͳͲɺ௕ظతɾࠜຊతରॲ͸೉͔ͬͨ͠ • ʮࢦඪΛݩʹػೳཁٻͱඇػೳཁٻͷ༏ઌ౓൑அ͕Ͱ͖Δʯ·Ͱ౸ୡ͠ͳ͔ͬͨ • ༏ઌ౓൑அʹ໾ʹཱͨͳ͍ͷͰ͋Ε͹ɺ։ൃνʔϜʹͱͬͯ΍Δ͜ͱ͕૿͚͑ͨͩͱ΋ݴ͑Δ

Slide 38

Slide 38 text

ͳͥ”ߦಈ͢Δ”·ͰࢸΒͳ͔ͬͨͷ͔ • ৴པੑͷ໰୊(ඇػೳཁٻ)ʹରॲ͢Δݖݶɾ༧ࢉෆ଍ • ػೳ։ൃͷ༏ઌ౓͕ߴ͔ͬͨ

Slide 39

Slide 39 text

SLO ҧ൓ͨ͠Β“ϦϦʔεετοϓ”ͱ͍͏ݬ૝ • SLO ҧ൓ͨ͠Βશͯͷ։ൃΛࢭΊͯͦͷमਖ਼Λߦ͏ͷ͸ݱ࣮ తͰ͸ͳ͍ • ͦΕΛߦ͏ͨΊʹ͸ࣄۀ੹೚ऀ͕͜ͷӡ༻ʹ߹ҙ͢Δඞཁ͕͋Δ • ͦΕΛߦ͏ͨΊʹ͸”׬ᘳͳSLI/SLO”͕ඞཁ͕ͩɺ”׬ᘳͳSLI/SLO”͸ ଘࡏ͠ͳ͍ • ૊৫͕֦େ͠ɺBiz Dev / Product / CS ͱ੹຿Λ෼୲͍ͯ͠Δঢ়گͰɺ ݱ৔൑அͰ౎౓͜ΕΛ΍Δͷ͸ίϛϡχέʔγϣϯίετ͕ߴ͗͢Δ

Slide 40

Slide 40 text

SLI/SLO ӡ༻ͷཧ૝: ͱ͋Δ Developer ͷίϝϯτ

Slide 41

Slide 41 text

ʮSLO Reviewʯat SRE NEXT 2020 ͦͷޙͷ·ͱΊ • ʮ৴པੑࢦඪΛఆΊɺ؍࡯͢ΔʯจԽΛ࡞ͬͨ͜ͱʹ͸Ձ஋͕͋ͬͨ • ࣄۀઓ্ུͷҙࢥܾఆʹ໾ʹཱͭࢦඪʹҭͭ·Ͱʹ͸ࢸΒͳ͔ͬͨ • ཧ༝1. ඇػೳཁٻͱػೳཁٻͷόϥϯεΛม͑Δҙࢥܾఆݖݶɾ༧ࢉ͕ϓϩμΫτ ։ൃνʔϜʹͳ͔ͬͨ • ৽نػೳ։ൃͷΠϯηϯςΟϒ͕େ͖͍ঢ়گ • ͦͷΑ͏ͳٕज़ઓུ/ٕज़౤ࢿΛϓϩμΫτ All Ͱߦ͑Δ࢓૊Έ͕ͳ͔ͬͨ • ཧ༝2. ৴པੑࢦඪ͕ Biz/Dev/SRE શһ͕ཧղ͠΍͍͢ࢦඪͰ͸ͳ͔ͬͨ • backend API ͷ SLI ͸ϢʔβମݧΛ௚઀ද͓ͯ͠ΒͣɺLatency ʹؔ͢Δରॲ͸ TPM ΁ͷઆ໌΋೉͍͠

Slide 42

Slide 42 text

Agenda • લఏɿSRE Λ࣮ݱ͢Δͱ͸Ͳ͏͍͏͜ͱ͔ • ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱࣄۀʹ͓͚Δ SRE • 2020೥ʮSLO ReviewʯͰͷֶͼͱࣦഊ • SRE ͱٕज़ઓུ • ·ͱΊͱࠓޙ

Slide 43

Slide 43 text

2021೥ ελσΟαϓϦ খதߴϓϩμΫτ։ൃ෦ ٕज़ઓུάϧʔϓൃ଍

Slide 44

Slide 44 text

ελσΟαϓϦখֶɾதֶɾߴߍɾେֶडݧߨ࠲ ελσΟαϓϦ For TEACHERS ελσΟαϓϦ For SCHOOL ݱঢ়ͷ૊৫ਤ: খதߴϓϩμΫτ։ൃ෦ ҎԼ17άϧʔϓ TPM BtoB TPM BtoC TPM ForSCHOOL TPM ԣஅ BtoC BtoB QA ։ൃࢧԉ SRE ٕज़ઓུ ίʔνϯά ৽ن։ൃ1 Τϯϋϯε ֶशࢧԉ Native iOS Android ৽ن։ൃ2 ਐ࿏ओମੑ ίϛϡχέʔγϣϯࢧԉ ForSCHOOLϞόΠϧ

Slide 45

Slide 45 text

Disclaimer • ٕज़ઓུάϧʔϓͷ্ཱͪ͛͸લ೚Ϛωʔδϟ͕ߦͬͨ΋ͷ • ࡢ೥࣌఺Ͱ͸ @chaspy ͸ DevOps WG ͷ Lead -> EM/Lead • લ೚ͷୀ৬ʹ൐͍ٕज़ઓུάϧʔϓͷ EM ͸෦௕͕݉೚ͭ͠ ͭɺଞ਺໊ͷ EM ͱҰॹʹӡӦ͍ͯ͠Δ • SLO ҧ൓ͷରॲ͕Ͱ͖ͳ͍͜ͱ͕ཧ༝Ͱ্ཱ͕ͪͬͨΘ͚Ͱ ͸ͳ͍

Slide 46

Slide 46 text

ٕज़ઓུͱ͸ ϓϩμΫτ։ൃ૊৫ʹ͓͍ͯ ԿΛ΍Δ͔ɺԿΛ΍Βͳ͍͔

Slide 47

Slide 47 text

ͳٕͥज़ઓུ͕ඞཁ͔ • Ϗδωεͷ଎౓ʹ௥ਵ͢ΔͨΊ • มߋΛ͠΍͍͢γεςϜɾίʔυɾ૊৫ʹ͢Δ • มߋ଎౓Λ๦͛ΔཁҼ͸ઌճΓͯ͠ରॲ͢Δ(ٕज़తෛ࠴ͷฦࡁ) • ͦͷঢ়ଶΛ௕ظతʹҡ࣋Ͱ͖ΔೳྗΛಘΔ

Slide 48

Slide 48 text

ͳٕͥज़ઓུ”άϧʔϓ”͕ඞཁ͔ • ٕज़ઓུͷܾΊํ͸૊৫ʹΑͬͯҟͳΔ • 1ਓͷ CTO ͕τοϓμ΢ϯͰܾΊͯ΋͍͍ • ϘτϜΞοϓͰશһ߹ٞͰܾΊͯ΋͍͍ • ͦͷதؒͰ΋͍͍ • ελσΟαϓϦখதߴ։ൃ૊৫͸ٕज़ઓུΛ1ਓʹґଘ͠ͳ͍࢓૊ΈΛ ࡞Δ͜ͱʹ௅ઓ͍ͯ͠Δ

Slide 49

Slide 49 text

ٕज़ઓུάϧʔϓͱ͸ • ໨త • ϓϩμΫτ։ൃ૊৫ͱͦͷγεςϜΛΑΓมԽʹڧ͘͢Δ • ໨ඪ • ٕज़తͳϏδϣϯͱํ਑ͷࡦఆ • ٕज़త՝୊ɾෛ࠴ΛίϯτϩʔϧԼʹஔ͘ • վળαΠΫϧͷཱ֬ͱࣗݾ਍அೳྗͷ֫ಘ

Slide 50

Slide 50 text

࠷ॳͷ1೥Ͱ΍ͬͨ͜ͱ • ৽ن։ൃɿΤϯϋϯεɿٕज़తෛ࠴ղফ = 1:1:1 ͷ༧ࢉએݴ • ٕज़తෛ࠴ղফҊ݅ΛܾΊͯ։ൃϩʔυϚοϓʹ૊ΈࠐΉ • ٕज़՝୊ͷ؅ཧํ๏ͱൃੜ࣌ͷτϦΞʔδํ๏ͷܾఆ

Slide 51

Slide 51 text

SLO ҧ൓࣌ͷΞΫγϣϯ͕औΕΔΑ͏ʹ࢓૊Έ͕Ͱ͖ͨ • Ұఆඇػೳཁٻʹ༧ࢉ͕౰ͯΒΕɺൃੜٕͨ͠ज़՝୊Λ؅ཧ ͠ɺ࣮ߦͰ͖Δମ੍͕੔ͬͨ • SLO ҧ൓ͨ͠ࡍʹ࣮ࢪ͢Δ͜ͱ͕೉͔ٕͬͨ͠ज़՝୊Λܭը ʹ૊ΈࠐΊΔΑ͏ʹͳͬͨ

Slide 52

Slide 52 text

׆ಈମ • ໨త • ϓϩμΫτ։ൃ૊৫ͱͦͷγεςϜΛΑΓมԽʹڧ͘͢Δ • ໨ඪ • ٕज़తͳϏδϣϯͱํ਑ͷࡦఆ • ٕज़త՝୊ɾෛ࠴ΛίϯτϩʔϧԼʹஔ͘ • վળαΠΫϧͷཱ֬ͱࣗݾ਍அೳྗͷ֫ಘ DevOps WG ԣஅWG Backend WG Frontend WG

Slide 53

Slide 53 text

Slide 54

Slide 54 text

ٕज़త՝୊ɾෛ࠴ΛίϯτϩʔϧԼʹஔ͘ • ٕज़՝୊Λ౤͛ࠐΊΔ৔ॴɾ࢓૊ΈΛ࡞Δ • ٕज़՝୊ͷ༏ઌ౓Λ͚ͭΔ • དྷظͷ։ൃϩʔυϚοϓʹ઀ଓ͢Δ

Slide 55

Slide 55 text

ٕज़՝୊Λ౤͛ࠐΊΔ৔ॴɾ࢓૊ΈΛ࡞Δ ୭Ͱ΋ٕज़՝୊Λ౤͛ࠐΊΔ spreadsheet reacji Ͱؾܰʹ౤͛ࠐΊΔ

Slide 56

Slide 56 text

౤͛ࠐ·Εͨ՝୊ͷ re fi nement ͭΒ͞ = ϖΠϯΛݴޠԽ ݱ࣌఺Ͱු͔Ϳ How Λهࡌ (༏ઌ౓͕͚ͭΒΕΕ͹͍͍ͷͰɺ͜ͷཝ͸ optional)

Slide 57

Slide 57 text

՝୊ͷ༏ઌॱҐ෇͚ ϖΠϯͷස౓ͱڧ౓ͰϚοϐϯά ٕज़՝୊ͷ૬ରධՁΛ͚ͭΔ

Slide 58

Slide 58 text

དྷظͷ։ൃϩʔυϚοϓʹ઀ଓ͢Δ ϖΠϯͷ਺஋Խ ։ൃϩʔυϚοϓ઀ଓঢ়گ

Slide 59

Slide 59 text

΋ͪΖΜɺ׬ᘳͰ͸ͳ͍ • ٕज़՝୊͸ස౓ͱڧ౓ͰଌΕΔ΋ͷͰ͸ͳ͍ • ఆੑతͰ͋Δ • ࢀՃϝϯόʔͷภΓ͕͋Δ͔΋ • ෳ਺ member ͷ vote ݁Ռͷॏ৺ʹஔ͍͍ͯΔͷͰਫ਼౓ʹٙ໰ • ։ൃϦιʔεɺٕज़త೉қ౓ɺϦεΫʹΑ͙ͬͯ͢ʹऔΓ͔͔Εͳ ͍՝୊΋͋Δ • ՝୊ͷ༏ઌ౓෇͚ʹ͕͔͔࣌ؒΔ • etc…

Slide 60

Slide 60 text

͔͠͠ɺಛఆͷਓʹґଘ͠ͳ͍ελʔτϥΠϯʹཱͬͨ • ୭Ͱ΋ٕज़՝୊Λද໌͢Δ͜ͱ͕Ͱ͖Δ • ද໌ͨ͠՝୊͸ඞͣऔΓѻΘΕΔ • શһͰ໰୊ʹ޲͖߹͑Δ - ໰୊ vs ࢲͨͪ • ௕ظతͳ৴པੑͷ Controlability ୲อʹ໾ཱͭ

Slide 61

Slide 61 text

Slide 62

Slide 62 text

DevOps WG ͷ໨తͱ׆ಈ • ໨తɿʢ։ൃνʔϜͷʣࣗݾ਍அೳྗͷ֫ಘͷͨΊʹઃஔ • ϝϯόʔ͸ ྖҬ͝ͱͷ WebDev / QA / SRE • ׆ಈ಺༰ • όϦϡʔετϦʔϜϚοϐϯάͷ࣮ࢪ • ީิͱͳΔ Metrics / Indicator ͷચ͍ग़͠ͱܭଌ • DX Criteria ͷ࣮ࢪ • όϦϡʔετϦʔϜΛ્֐͢ΔཁҼͷղܾ(e.g. E2E Automation) • ϓϩμΫτ։ൃ෦֎΁ͷ޿ใ׆ಈ • ·ͣ͸༗ޮͦ͏ͳ metrics ΍ΞηεϝϯτΛݕূͨ͠

Slide 63

Slide 63 text

όϦϡʔετϦʔϜϚοϐϯά Meta Issue (Epic) ࡞੒~ ։ൃ

Slide 64

Slide 64 text

όϦϡʔετϦʔϜϚοϐϯά ։ൃ׬ྃ ~ εϓϦϯτϨϏϡʔ QA ~ ຊ൪ϦϦʔε

Slide 65

Slide 65 text

ީิͱͳΔ Metrics / Indicator ͷચ͍ग़͠ͱܭଌ όϦϡʔετϦʔϜϚοϐ ϯάΛΠϯϓοτʹɺվ ળͨ͠৔߹ͷϝϦοτͷ Ծઆ΍ɺͦΕΛ્֐ͯ͠ ͍Δ໰୊ΛϐοΫΞοϓ

Slide 66

Slide 66 text

Q https://blog.studysapuri.jp/entry/2020/08/17/dx-criteria-system

Slide 67

Slide 67 text

ϓϩμΫτ։ൃ෦֎Ͱͷ޿ใ׆ಈ: BtoC All Hands Ͱͷൃද https://blog.studysapuri.jp/entry/2020/08/17/dx-criteria-system • ॴଐάϧʔϓΛ௒͑ͨࣄۀঢ়گΛ஌Δ৔ • Ϛʔέοτχϡʔε • ࣄۀঢ়گ • ϓϩμΫτ KPI • SLI / ։ൃऀੜ࢈ੑ • ͦͷଞτϐοΫ͞·͟· • SRE ͱ͸ͳʹʁ • ϚΠΫϩαʔϏεͬͯͳʹʁ͏Ε͍͠ͷʁ

Slide 68

Slide 68 text

ϓϩμΫτ։ൃ෦֎Ͱͷ޿ใ׆ಈ: BtoC All Hands Ͱͷൃද • “͸͡Ίͯͷ SRE” ͱ୊ͯ͠։ൃ෦Ҏ֎ͷํ޲͚ʹൃද

Slide 69

Slide 69 text

ϓϩμΫτ։ൃ෦֎Ͱͷ޿ใ׆ಈ: BtoC All Hands Ͱͷൃද • “͸͡Ίͯͷ SRE” ͱ୊ͯ͠։ൃ෦Ҏ֎ͷํ޲͚ʹൃද

Slide 70

Slide 70 text

DevOps WG ͷ໨తͱ׆ಈ • ໨తɿʢ։ൃνʔϜͷʣࣗݾ਍அೳྗͷ֫ಘͷͨΊʹઃஔ • ʢ։ൃνʔϜ͕ʣࢦඪΛݩʹߦಈ͢ΔจԽΛ࡞Ζ͏ͱ͍ͯ͠Δ • SLI/SLO Λݟͯ৴པੑΛ؍࡯͍ͯ͘͜͠ͱͱಉ͡Ͱ͸…?

Slide 71

Slide 71 text

SRE ͱٕज़ઓུ • DevOps WG ͷ׆ಈ͸ ʮSRE ͷ࣮ݱʯͷจԽ໘Ͱͷ֦ு • զʑ͕ݟΔ΂͖ࢦඪ͸γεςϜͷ৴པੑࢦඪ͚ͩͰ͸ͳ͍ • ͋ΒΏΔ΋ͷΛࢦඪΛݟͯɺҙࢥܾఆ͢Δ • ࠓޙ͸͜ͷจԽৢ੒ͦͷ΋ͷͷվળαΠΫϧΛճ͢ • 1. ީิͱͳΔ metrics ͷ༗ޮੑ͕໌Β͔ʹͳΓɺ਺஋Խ͢Δ • 2. ։ൃνʔϜ͕ͦΕΛݟͯɺΞΫγϣϯΛߟ͑Δ͜ͱ͕Ͱ͖Δ • 3. ։ൃνʔϜ͕ΞΫγϣϯ->վળͷαΠΫϧΛճ͢ • 4. 1-3 ͦΕࣗମ͕͏·͍͍ͬͯ͘Δ͔ΛධՁ͢Δ

Slide 72

Slide 72 text

SRE ͱٕज़ઓུ: ·ͱΊ • SRE Λ࣮ݱ͢ΔͨΊʹ͸ɺSLO ҧ൓Λͨ࣌͠ʹߦಈͰ͖Δ༧ࢉͱݖݶ ͕ඞཁ • ͦͷ্Ͱɺٕज़՝୊Λղܾ͢Δ༏ઌॱҐΛ͚ͭΒΕΔٕज़ઓུ͕ඞཁ • ʰελσΟαϓϦʱখதߴϓϩμΫτ։ൃ෦Ͱ͸͜ͷٕज़ઓུΛ1ਓ ʹґଘͤͣɺάϧʔϓͰ࣮ݱ͢Δ͜ͱʹ௅ઓ͍ͯ͠Δ • ͋ΒΏΔ΋ͷΛࢦඪͰݟ͍ͯ͘จԽ͕৴པੑͷͨΊʹॏཁ

Slide 73

Slide 73 text

Agenda • લఏɿSRE Λ࣮ݱ͢Δͱ͸Ͳ͏͍͏͜ͱ͔ • ྺ࢙͔ΒৼΓฦΔʰελσΟαϓϦʱࣄۀʹ͓͚Δ SRE • 2020೥ʮSLO ReviewʯͰͷֶͼͱࣦഊ • SRE ͱٕज़ઓུ • ·ͱΊͱࠓޙ

Slide 74

Slide 74 text

·ͱΊ / SRE Λ࣮ݱ͢ΔͨΊʹඞཁͳ͜ͱ • ࢦඪΛݩʹվળ͠ଓ͚ΔจԽ • ࣗݾ׬݁ԽΛࢧ͑Δ Platform • ༏ઌॱҐΛมߋ͢ΔͨΊͷٕज़ઓུ

Slide 75

Slide 75 text

SRE Λ࣮ݱ͢Δͱ͸

Slide 76

Slide 76 text

։ൃνʔϜ͕৴པੑΛ ίϯτϩʔϧ͢Δ Capability Λ ਎ʹ͚͍ͭͯΔ͜ͱ

Slide 77

Slide 77 text

SRE “NEXT” in ʰελσΟαϓϦʱ • “৴པੑ” ʹؔͯ͠͸ Enabling Team ͱͯ͠ͷ SRE Team ͸໾ ׂΛՌͨͭͭ͋͠Δ • SRE Team ͷࠓޙ • ΑΓ૊৫Λ Sustainable / Scalable ʹ͢ΔͨΊʹɺPlatform ʹؔ͢Δ ΦϯϘʔσΟϯάͷ֦ॆ΍ɺ։ൃνʔϜ͕ࣗ཯తʹ৴པੑʹؔ͢Δ Capability शಘΛ൑அͰ͖ΔΞηεϝϯτΛఏڙ͢Δ • ৴པੑ͚ͩͰ͸ͳ͍ɺ։ൃੜ࢈ੑΛՌͨͤΔ Platform ։ൃʹ஫ྗ͢Δ

Slide 78

Slide 78 text

Who owns the Service Level? • Service Level ͸ϓϩμΫτʹؔΘΔશһͷ΋ͷ • શһ͕ؔ৺Λ࣋ͯΔΑ͏ͳ৴པੑࢦඪʹਐԽͤ͞·͠ΐ͏ • ϢʔβମݧΛ௚઀తʹද͢ Client-side(WebFrontend/Native) Ͱͷ SLI/SLO Λ௥͏ • ʮࢦඪΛݟͯߦಈ͢Δʯͦͷ΋ͷͷվળαΠΫϧΛճ͠·͠ΐ͏

Slide 79

Slide 79 text

Thank you! chaspy chaspy_ Engineering Manager, Site Reliability at Recruit Co., Ltd. Takeshi Kondo https://chaspy.me