Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Sign up for free
Menu
Search
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Features
All features
Private URLs
Password Protection
Custom URLS
Scheduled publishing
Remove Branding
Restrict embedding
Deck Collections
Notes
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Explore
Featured decks
Featured speakers
Programming
Technology
Storyboards
Pricing
Search
Sign in
Sign up for free
TimeFuzeアーキテクチャ構想 / The concept of TimeFuze Arc...
Search
Yuuki Tsubouchi (yuuk1)
December 23, 2017
Research
10k
1
Share
Embed
Copy iframe code
Copy JS code
Copy link
Start on current slide
TimeFuzeアーキテクチャ構想 / The concept of TimeFuze Architecture
第1回ウェブシステムアーキテクチャ(WSA)研究会@京都
Yuuki Tsubouchi (yuuk1)
December 23, 2017
More Decks by Yuuki Tsubouchi (yuuk1)
See All by Yuuki Tsubouchi (yuuk1)
「アラーティング」の話をしよう— SREconや論文等の最先端とのギャップをみる
yuukit
7
2.1k
SAKURAONE:An Open Ethernet-based AI HPC System And Its Observed Workload Dynamicsin a Single-Tenant LLM Development Environment
yuukit
1
660
AIスーパーコンピュータにおけるLLM学習処理性能の計測と可観測性 / AI Supercomputer LLM Benchmarking and Observability
yuukit
1
1.2k
SREはサイバネティクスの夢をみるか? / Do SREs Dream of Cybernetics?
yuukit
3
850
SREのためのテレメトリー技術の探究 / Telemetry for SRE
yuukit
14
3.9k
AIスパコン「さくらONE」の オブザーバビリティ / Observability for AI Supercomputer SAKURAONE
yuukit
2
1.5k
AIスパコン「さくらONE」のLLM学習ベンチマークによる性能評価 / SAKURAONE LLM Training Benchmarking
yuukit
2
1.3k
とあるSREの博士「過程」 / A Certain SRE’s Ph.D. Journey
yuukit
11
7.7k
eBPFを用いたAIネットワーク監視システム論文の実装 / eBPF Japan Meetup #4
yuukit
3
1.9k
Other Decks in Research
See All in Research
モデル評価を効率的にする方法
ftakahashi
0
190
[SNLP2026] Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach
wataruuuuu
0
330
MIRU2026 チュートリアル講演2:三次元データ処理の動向
nnchiba
8
5.3k
Fukui Shibiten 39 - AI Art
butchi
0
260
データサイエンティストの就労意識~2015 → 2026 一般(個人)会員アンケートより
datascientistsociety
PRO
0
990
La génomique au service de la fromageabilité du lait grâce aux spectres MIR
institutdelelevage
PRO
0
160
[CV勉強会@関東 CVPR2026] PSDesigner: Automated Graphic Design with a Human-Like Creative Workflow / kantocv 67th CVPR 2026
shunk031
0
360
某助成金プロジェクト採択に向けて企業研究所のアウトリーチ専任者がやったこと
afroscript
0
190
CVPR2026論文紹介_VLMにとって良いvision encoderとは何か?Rethinking Model Selection in VLM Through the Lens of Gromov-Wasserstein Distance
kobayashi31
1
260
COMETAを用いたデータ民主化運動の歴史
sazimai
0
270
IA for theory
gpeyre
1
470
人間中心の意思決定支援AI
yukinobaba
PRO
7
4.2k
Featured
See All Featured
Taking LLMs out of the black box: A practical guide to human-in-the-loop distillation
inesmontani
PRO
3
2.4k
Optimizing for Happiness
mojombo
378
71k
Building a Modern Day E-commerce SEO Strategy
aleyda
45
9.2k
End of SEO as We Know It (SMX Advanced Version)
ipullrank
3
4.4k
The Spectacular Lies of Maps
axbom
PRO
1
1.1k
Un-Boring Meetings
codingconduct
0
440
sira's awesome portfolio website redesign presentation
elsirapls
0
440
No one is an island. Learnings from fostering a developers community.
thoeni
21
3.8k
Understanding Cognitive Biases in Performance Measurement
bluesmoon
32
3k
RailsConf 2023
tenderlove
30
1.6k
[RailsConf 2023 Opening Keynote] The Magic of Rails
eileencodes
31
10k
Measuring Dark Social's Impact On Conversion and Attribution
stephenakadiri
2
300
Transcript
TimeFuzeΞʔΩςΫνϟߏ ͯͳ id:y_uuki / @y_uuk1 2017/12/23 ୈ1ճΣϒγεςϜΞʔΩςΫνϟ(WSA)ݚڀձ@ژ ॲཧͱσʔλͱλΠϚʔΛҰମԽͨ͠σʔλύΠϓϥΠϯ
id:y_uuki / @y_uuk1 Yuuki TSUBOUCHI https://yuuk.io/ ͯͳͷΣϒΦϖϨʔγϣϯΤϯδχΞ WSAݚڀձ ཧࣄ?෭ࠪ?
TimeFuze?
࣌ݶ৴ Time Fuze
ىരஔ+Րༀ+λΠϚʔ
ॲཧ+σʔλ+λΠϚʔ
TimeFuze ΞʔΩςΫνϟ
͘͡ 1. എܠͱత 2. TimeFuzeΞʔΩςΫνϟ 3. ߟ 4. Ή͢ͼ
1. എܠͱత
WebαʔϏεͷσʔλॲཧཁٻͷෳࡶԽ • ϦΫΤετʹରͯ͠HTMLͷϨεϙϯεΛฦ٫͢ΔҎ֎ ͷϫʔΫϩʔυͷଟ༷Խ͕ਐΜͰ͍Δ • ੑೳಛੑͷҟͳΔσʔλετΞͷซ༻ɺػցֶशʹΑ Δ࣌ؒपظͰͷେنͳσʔλॲཧΛٻΊΒΕ͍ͯΔ • εέʔϧ͍ͤ͢͞σʔλύΠϓϥΠϯॲཧ͕ඞཁ
ҰൠͷσʔλύΠϓϥΠϯ DataSource Data Destination όονॲཧ λΠϜεέδϡʔϥʔ ىಈ
ҰൠͷσʔλύΠϓϥΠϯͷ՝ • σʔλྔ૿େʹ͋ΘͤͨεέʔϦϯάӡ༻ͷ͠͞ • όονॲཧ్தͷΤϥʔॲཧͷ͠͞ • λΠϜεέδϡʔϥͦͷͷͷӡ༻ͷࡶ͞ • ࣮ߦසઃఆͷॊೈੑͷ͞
σʔλ λΠϚʔ σʔλͱॲཧͱλΠϚʔͷҰମԽϞσϧͷఏҊ ॲཧ DataSource σʔλ λΠϚʔ ॲཧ DataDestination
ҰൠͷσʔλύΠϓϥΠϯͷ՝ͷղܾ • σʔλྔ૿େʹ͋ΘͤͨεέʔϦϯάӡ༻ͷ͠͞ • ࠷ॳ͔ΒฒߦॲཧΛલఏͱͨ͠ΞʔΩςΫνϟͰղܾ • όονॲཧ్தͷΤϥʔॲཧͷ͠͞ • όονॲཧͰͳ͘Ϩίʔυ୯ҐͰδϣϒఆٛͰղܾ •
λΠϜεέδϡʔϥͦͷͷͷӡ༻ͷࡶ͞ • σʔλετΞͷTTLʹΑΓεέδϡʔϥϗετ͕ෆཁ • ࣮ߦසઃఆͷॊೈੑͷ͞ • λΠϚʔઃఆͷཻΛϨίʔυ୯ҐͰઃఆ͢Δ͜ͱͰղܾ
http://blog.yuuk.io/entry/the-rebuild-of-tsdb-on-cloud
֊ܕσʔλετΞߏͷ࣌ܥྻσʔλϕʔε ΠϯϝϞϦ DB ΦϯσΟεΫ DB ΦϒδΣΫτ ετϨʔδ ৽ ݹ ߴ
҆
֊ܕσʔλετΞͷ͍͠ͱ͜Ζͱղܾ • σʔλετΞؒͷσʔλҠಈύΠϓϥΠϯͷεέʔϥϏ ϦςΟ • ಛఆͷϨίʔυͷΈɺσʔλετΞҠಈλΠϛϯάΛ੍ ޚ͢Δ • ͜ΕΒΛTimeFuzeΞʔΩςΫνϟʹΑΓղܾ •
DynamoDBͷTTLͱLambda TriggerʹΑΔ࣮
ͦͷଞͷTimeFuzeͷద༻ΞϓϦέʔγϣϯྫ • ࣌ܥྻσʔλҟৗݕͷֶशϞσϧͷఆظߋ৽ • Ϛϧνςφϯτڥʹ͓͚ΔେنͳSSL/TLSূ໌ॻߋ ৽
2. TimeFuzeΞʔΩςΫνϟ
1. insert DataSource record Data TTL DataDestination Trigger Other API/DataStore
record 2. expire 3. insert or update
TimeFuzeͷϝϦοτ • ฒߦॲཧΛલఏͱͨ͠ΞʔΩςΫνϟͰ͋Δ͜ͱ • Τϥʔ͔Βͷճ෮ॲཧ͕༰қͰ͋Δ͜ͱ • λΠϜεέδϡʔϥͱόονॲཧͷͨΊͷϗετ͘͠ Ϋϥελͷߏஙɾӡ༻͕ෆཁͰ͋Δ͜ͱ • Ϩίʔυ୯ҐλΠϚʔʹΑΓɺδϣϒͷ࣮ߦසΛϨίʔ
υ୯ҐͰௐՄೳͰ͋Δ͜ͱ
AWSʹΑΔ࣮
DynamoDB TTL + Lambda Trigger • DynamoDB: ϑϧϚωʔδυܕͷNoSQLσʔλϕʔε αʔϏεɻTTLΛαϙʔτ͍ͯ͠Δɻ •
Lambda: ΠϕϯτൃՐΛܖػʹFunctionΛݺͼग़ͤΔ • DynamoDB্ͷϨίʔυʹର͢ΔTTLظݶΕΠϕϯτ ΛܖػʹɺLambda FunctionΛτϦΨʔͱ࣮ͯ͠ߦ
1. insert DynamoDB record Data TTL DynamoDB,S3,… Lambda Other API/DataStore
record 2. expire 3. insert or update
ΦϯϓϨϛεڥͰͷ࣮
Redis Keyspace௨ + ϝοηʔδΩϡʔ • RedisͷKeyspace௨ػೳʹΑΓTTLΠϕϯτΛߪಡͰ͖ Δ • ͨͩ͠ɺ҆શͳτϦΨʔॲཧͷͨΊʹɺߪಡͨ͠Πϕϯ τΛϝοηʔδΩϡʔʹΑΓӬଓԽ͢Δඞཁ͕͋Δ
• ݴޠΘͣར༻͍͢͠ϝοηʔδΩϡʔͱͯ͠ Fireworq͕͋Δ
TimeFuzeͷ ۩ମతͳΞϓϦέʔγϣϯ ͷద༻
࣌ܥྻσʔλͷҟৗݕ • ऩूͨ͠ϝτϦοΫܥྻʹରͯ͠ɺֶशϞσϧΛߏங • ৽نϝτϦοΫͷରԠͷͨΊϞσϧΛఆظతʹߋ৽͢Δ • ଞͷσʔλετΞʹҠಈ͢ΔΘ͚Ͱͳ͘ɺDataSource ͱDataDestinationಉҰͷσʔλετΞΛར༻ • ֶ֤शϞσϧͷϨίʔυʹTTLΛઃఆ͠ɺTriggerͰֶश
ॲཧޙɺϞσϧΛߋ৽͢Δ
1. insert DataSource record Metric TTL Trigger ࣌ܥྻ σʔλϕʔε 2.
expire 3. update ֶशॲཧ
େنSSL/TLSূ໌ॻཧ • ϒϩάαʔϏεͳͲͰେྔͷSSL/TLSূ໌ॻΛཧ͠ɺ ϲ݄͝ͱʹߋ৽͢Δέʔε • ূ໌ॻΛσʔλϕʔεʹ֨ೲ͠ಈతऔಘ͢ΔΞʔΩςΫ νϟΛલఏͱ͢Δ • ֤υϝΠϯʹର͢Δূ໌ॻϨίʔυʹTTLΛઃఆ͠ɺ Trigger͕ূ໌ॻΛ࠶औಘ͠ɺDataDestination্ॻ͖
1. insert DataSource record ূ໌ॻ TTL Trigger Let’s Encrypt 2.
expire 3. update ূ໌ॻ ߋ৽ॲཧ
3. ߟ
TimeFuzeͷσϝϦοτ • ΞʔΩςΫνϟϨϕϧ • DataSource্ͷෳͷϨίʔυΛϚʔδ͢ΔॲཧΛͮ͠Β͍ • ࣮Ϩϕϧ • Ұ࣌తʹϨίʔυ͕আ͞ΕΔͨΊɺσʔλͷҰ؏ੑͷ ͕ඞཁ
• TTLظݶ͔Βআॲཧ։࢝·ͰͷσΟϨΠ͕࣌ؒϕετΤ ϑΥʔτͳͨΊɺִ࣌ؒؒ
TimeFuzeͷσϝϦοτղܾҊ • TTLΛར༻ͤͣɺTime to EventػೳΛσʔλετΞʹ ΈࠐΉ • Time to Eventػೳઃఆͨ࣌͠ࠁΛܦաͨ͠Πϕϯτ
ͷΈΛൃߦ͠ɺϨίʔυআ͠ͳ͍
4. Ή͢ͼ
Ή͢ͼ • εέʔϧ͍ͤ͢͞σʔλύΠϓϥΠϯΞʔΩςΫνϟ͕ WebαʔϏε։ൃʹٻΊΒΕ͍ͯΔ • ୯ҰͷλΠϜεέδϡʔϥʔʹΑΔσʔλύΠϓϥΠϯͷ ՝ΛɺσʔλͱॲཧͱλΠϚʔͷҰମԽʹΑΓղܾ͢Δ ख๏ΛఏҊͨ͠ • ࣮ͷʹΑΓɺݱஈ֊Ͱ࣮ӡ༻͢ΔͨΊʹΞϓϦέʔ
γϣϯͷ͕ඞཁͱͳΔ
ࠓޙͷऔΓΈ • طଘͷఆظσʔλύΠϓϥΠϯͷ՝ΛαʔϕΠ͠ཧ • طଘσʔλετΞ࣮ͷTTLσΟϨΠ࣌ؒͷௐࠪ • طଘͷσʔλετΞͷTime to EventͷΈࠐΈ
ࣗੑͷ͋ΔσʔλॲཧͷՄೳੑ • VM => ίϯςφ, ίϯςφ => εϨου • σʔλϕʔε
=> ςʔϒϧ, ςʔϒϧ => σʔλϨίʔυ
SRE/ΦϖϨʔγϣϯΤϯδχΞͷઓ • TimeFuzeͷΑ͏ͳσʔλॲཧΞʔΩςΫνϟΛࣄʹ͍ͯ͠ͳ͍ • ۙͰ͍ͬͯΔ͜ͱɺΠϯϑϥετϥΫνϟࣗମͷมߋΛ࠷ େԽ͢Δࣄ • ج൫ͷந͕͕͋ΔʹͭΕɺSREͷࣄ͕ΑΓࣄۀʹۙ͘ͳ͍ͬͯ͘ • ࣄۀʹ͍ۙ΄ͲɺநԽ͕͘͠ͳΓج൫Ͱٵऩͮ͠Β͘ͳΔ
• ෳͷσʔλετΞΛ͏·͘׆༻ͨ͠ΓɺෳͷϫʔΫϩʔυΛ͏· ͍͍ͯ͘͘͞ΞʔΩςΫνϟʹؔ৺͕͋Δ