Share
Embed
国際会議 IEEE SERVICES/CLOUD 2020 参加報告 / A Report of IEEE SERVICES/CLOUD 2020
クラウドコンピューティングを中心とした国際会議IEEE CLOUD 2020に参加してきたので、その報告をまとめています。
https://conferences.computer.org/cloud/2020/program/
Yuuki Tsubouchi (yuuk1)
October 28, 2020
Transcript
31 slides
-
2 1. IEEE SERVICES/CLOUD 2020ͷ֓ཁ 2. Best Paper Awardडจͷհ 3.
ௌߨͨ͠ݚڀൃදͷհ 4. ΦϯϥΠϯձͷ༷ࢠͱޮՌతͳௌߨͷͨΊͷ 5. ૯ׅͱདྷͷ։࠵༧ࠂ ࣍ -
4 IEEE SERVICES 2020 ɾ202010݄18-24 ژ → ΦϯϥΠϯ։࠵ ɾݩདྷ7݄։࠵͕ͩͬͨCOVID-19ͷӨڹʹΑΓԆظ͞ΕͨͷͪʹΦϯ ϥΠϯ։࠵ͱͳͬͨ
ɾIEEE Computer Society Technical Committee on Services Computing (TC-SVC) ओ࠵ ɾαʔϏείϯϐϡʔςΟϯάʹؔ࿈͢Δෳͷࠃࡍձٞͷڞ࠵Πϕϯτ ɾ2004͔Β։࠵͞Ε͍ͯΔ -
5 IEEE SERVICES 2020ͷ֤ࠃࡍձٞ ࠾ Core Rank IEEE CLOUD
ΫϥυίϯϐϡʔςΟϯά IaaSɺPaaSͳͲ 53/256 = 20.7% (regular) 13 papers (short) B IEEE ICWS WebϕʔεͷαʔϏεͷϥΠϑαΠ Ϋϧݚڀͱཧ 58/321 = 19% (regular) 20 papers (short) A IEEE SCC αʔϏείϯϐϡʔςΟϯά 41/200 = 20.5% (regular) 11 papers (workshop) 14 papers (WIP) A IEEE SMDS σʔλۦಈܕΞϓϦέʔγϣϯ σʔλੳɺσʔλ࣭ No data - IEEE EDGE ΤοδίϯϐϡʔςΟϯά 23/61 = 37.7% (regular) - CLOUD Λத৺ʹௌߨ -
6 Keynotes (1) On Odor Reproduction, and How to Test
For It by David Harel, a Vice President of the Israel Academy of Sciences and Humanities (2) Cloud Native Database Systems for Enterprise Applications by Feifei Li, a Vice President of Alibaba Group (3) On the Future of Smart Transportation with AI by Ozan Tonguz, a professor at Carnegie Mellon University (4) Digital Health and Beyond: What Computing Can Do to Transform Health by Wendy Nilsen, a Program Director in the Computer and Information Science and Engineering Directorate at NSF (5) IT Services after COVID-19 by Dr. Dong Xie, a CTO of IBM Greater China Group and Director of IBM China Research Lab (6) Participatory Systems in the Digital Era: A Services Computing Perspective by Valérie Issarny, Director of Research at INRIA. -
8 IEEE CLOUD 2020ͷ Cloud as a Service • IaaS,
PaaS, and SaaS • Function as a Service • Network as a Service • Storage as a Service • Everything as a Service Cloud Infrastructure • Cloud Computing System & Architectures • Edge Computing System & Architectures • Cloud-centric Network Architectures • Storage & Data Architectures • Hybrid-clouds & Multi-clouds Integration Cloud Applications • Large Scale Cloud Applications • Terminal-Edge-Cloud Applications • 5G/6G Enhanced Edge/Cloud Applications • Social & Mobile Cloud Applications • Innovative Cloud Applications Cloud Management and Operations • Distributed & Parallel Query Processing • Resource, Energy & Data Management • Cloud Metering & Monitoring • Containers & Serverless Computing • SDN, NFV, & Data Center Network • Cloud Service Adaptation & Automation • Cloud Federation & Service Composition Cloud Trustworthiness • Access Control, Authorization, & Authentication • Assurance, Audit, Certification, Compliance • Fault Tolerance, High Availability, & Reliability • Cryptographic Algorithms and Protocols • Cloud Security and Privacy • Trusted Cloud Environments https://conferences.computer.org/cloud/2020/cfp/ ΑΓҾ༻ ӡ༻ཧ ηΩϡϦςΟ ΞϓϦέʔγϣϯ ΠϯϑϥͷΞʔΩςΫνϟ αʔϏεͱͯ͠ͷΫϥυ -
9 IEEE CLOUD 2020ͷηογϣϯฤ ɾService Prediction & Optimization ɾEdge Applications
x 2 ɾEdge Caching & Computation Offloading ɾEdge-Cloud Collaboration ɾCloud & Edge Management ɾCloud Applications x 4 ɾCloud Management x 2 ɾCloud Management & Governance ɾSDN and NFV ɾNetwork and Storage Optimization ɾDeep Learning & Federated Learning ɾCloud Modeling x 2 ɾCloud Performance & Maintenance ɾCloud Security ɾCloud Task Scheduling ɾCloud Performance Evaluation ɾServerless Computers and Containers ɾMicroservices & Containers ɾOrchestration & Data Analysis ΤοδίϯϐϡʔςΟϯάɺਂֶशɺαʔ όϨε&ίϯςφͳͲ͕ϗοτͳτϐοΫ -
10 ɾژ։࠵ͳͷͰɺதࠃͷେֶʹॴଐ͢ΔํʑʹΑΔൃදཱ͕ͬͨ ɾͦͷଞɺUSAɺϤʔϩούݍɺΠϯυɺΦʔετϥϦΞͳͲ ɾຊࠃͷ৫ʹॴଐ͢ΔஶऀʹΑΔൃද3݅ ɾIBM Research - Tokyo x 1ɺभۀେֶ
x 2 ɾ֤ࠃͷIBM ResearchॴଐʹΑΔൃදཱ͕ͬͨ ɾΫϥυࣄۀऀॴଐͷൃද3݅ͱগͳΊ (Alibaba Cloud,IBM Cloud) ൃද͞ΕΔݚڀͷஶऀͷॴଐ -
11 ɾରγεςϜ: Ϋϥυ্ͷϚΠΫϩαʔϏεɺԠతʹεέʔϧ͢ ΔαʔόϨείϯϐϡʔςΟϯάɺଟͷΤοδͱΫϥυͷ࿈ܞͱ ͍ͬͨಈత͔ͭࢄͨ͠ෳࡶͳγεςϜߏ ɾҙࣝ: γεςϜཧऀʹΑΔϧʔϧϕʔεͷઃܭͷݶք ɾΞϓϩʔν: ͪߦྻཧཧ࠷దԽͳͲͷݹయతͳཧϞσ ϧͷద༻ʹՃ͑ͯɺػցֶशͱਂֶशΛద༻͢Δ
ɾ৽ͨʹML/DLΛద༻͢Δͱ͍͏ΑΓɺ͢ͰʹͦΕΒΛద༻ͨ͠ઌߦ ݚڀͷ՝Λղ͘ͱ͍͏ஈ֊ʹೖ͍ͬͯΔ ɾҎ্͕͜ͷखͷൃදͩͬͨΑ͏ʹײͨ͡ʢஶऀ؍ଌʣ Ұ෦ͷݚڀൃදʹڞ௨͢Δઃఆݚڀख๏ -
13 IEEE CLOUD 2020 Best Paper Award Skedulix: Hybrid Cloud
Scheduling for Cost-Efficient Execution of Serverless Applications Anirban Das (Rensselaer Polytechnic Institute), Andrew Leaf (Rensselaer Polytechnic Institute), Carlos A. Varela (Rensselaer Polytechnic Institute), Stacy Patterson (Rensselaer Polytechnic Institute) -
15 ɾϓϥΠϕʔτΫϥυͰOSSʹΑΓFaaSج൫ΛߏஙՄೳ ɾ͔͠͠ɺϓϥΠϕʔτΫϥυϦιʔε͕ݶΒΕ͍ͯΔͨΊɺϋΠ ϒϦουΫϥυΛ࠾༻͢Δ͜ͱ͕͋Δɻ ϓϥΠϕʔτΫϥυͰεέʔϧ͍ͤͨ͞ Private Cloud Public Cloud Incoming
jobs Additional jobs Function ɾϓϥΠϕʔτΫϥυͷ ϓϥΠόγʔɺੑೳɺίε τ໘ͰͷརΛڗडՄೳ ɾϐʔΫ࣌ͷෛՙ૿େʹඋ ͑ͨϋʔυΣΞͷҡ͕࣋ ෆཁ ίετ ߴεέʔϥϏϦςΟ -
16 ఏҊख๏ Skedulix ͷ֓ཁ ɾతϋΠϒϦουϓϥοτϑΥʔϜ্Ͱػೳ࣮ߦΛεέδϡʔϦϯ ά͠ɺύϒϦοΫΫϥυͷར༻ίετΛ࠷খݶʹ͑ͭͭɺϢʔβ ͕ࢦఆ࣮ͨ͠ߦ࣌ؒͷظݶΛຬͨ͢͜ͱ ɾ͜ͷεέδϡʔϦϯάͱׂΓͯͷܗࣜԽɺ ࠞ߹ઢܗ ͱ࣮ͯ͠ݱ͞ΕɺNPࠔͰ͋Δ͜ͱ͕ূ໌͞Ε͍ͯΔʢʣ
ɾSkedulixɺ࣮ؔߦ࣌ؒͱωοτϫʔΫԆͷ༧ଌϞσϧΛ༻͍ ͯɺίετ࠷খԽͷͨΊʹ֤ػೳͷॱংͱஔΛಈతʹܾఆ͢Δᩦཉ ͳΞϓϩʔνΛ࠾༻͢Δ ɾϓϥΠϕʔτΫϥυͷΈͱൺֱ͠ɺ40.5%ͷίετݮͱ࠷େ 1.92ഒͷߴԽΛୡͨ͠ -
εέδϡʔϥ 0. ճؼϞσϧʹΑΓ δϣϒͷ࣮ߦ࣌ؒΛ༧ ଌ͢ΔϞσϧΛߏங 1. όονδϣϒͷ ΩϡʔΠϯά 2. εέδϡʔϦϯάΞ
ϧΰϦζϜͷ࣮ߦ δϣϒͷ֤εςʔδ͝ ͱʹධՁ 18 εέδϡʔϥͷ֓ཁ -
21 Deep Unsupervised Workload Sequence Anomaly Detection with Fusion of
Spatial and Temporal Features in the Cloud Mengqing Wang, Zhihua Zhang, and others, IEEE CLOUD 2020. ࣗͷݚڀʹؔ࿈͢Δൃද Root-Cause Metric Location for Microservice Systems via Log Anomaly Detection Lingzhi Wang, Nengwen Zhao, and others, IEEE ICWS 2020. RAD: Detecting Performance Anomalies in Cloud-Based Web Services Joydeep Mukherjee, Alexandru Baluta, and others, IEEE CLOUD 2020. ҟৗݕ ҟৗݕ ҟৗݕͱ ݪҼಛఆ -
22 ɾखಈͰᮢΛઃఆෛՙγʔέϯεͷछྨ͕ଟ ͍ͱ͍ͨ͠ΊɺҟৗΛڭࢣͳ͠Ͱࣗಈݕग़ ɾෳͷෛՙγʔέϯεͷมԽύλʔϯʹదͨ͠ڭ ࢣͳ͠ͷҟৗγʔέϯεݕग़ϞσϧΛͲͷΑ͏ʹ ߏங͢Δ͔͕՝ ɾσΟʔϓχϡʔϥϧωοτϫʔΫΛ༻͍ͯɺγʔ έϯεͷຊ࣭తͳಛΛัଊ͠ɺҟৗγʔέϯε ͷແؔͳಛΛ͢Δ ɾۭؒతಛ->CNNɺ࣌ؒతಛ->BiLSTM
Deep Unsupervised Workload Sequence Anomaly Detection with Fusion of Spatial and Temporal Features in the Cloud -
23 ɾෳࡶԽͨ͠γεςϜͷҟৗͱͦ ͷݪҼΛࣗಈతʹݕग़͍ͨ͠ ɾ֎෦γεςϜͱͷڝ߹ͳͲɺγ εςϜԼҐͷϝτϦοΫ͚ͩͰ ҟৗͷݕग़͕ࠔ ɾ্ҐͷܭଌखؒΦʔό ϔου͕͔͔Δ RAD: Detecting
Performance Anomalies in Cloud-Based Web Services ɾVM͔ΒURLͱγεςϜϝτϦοΫΛܧଓతʹࢹ͠ɺΩϡʔΠϯά ωοτϫʔΫϞσϧΛߏங͢Δ ɾϞσϧͰ༧ଌͨ͠Ԡ࣌ؒΛɺҟৗ͕ͳ͍߹ͷϞσϧͰ༧ଌ͞Εͨ ϕʔεϥΠϯͱͳΔԠ࣌ؒͱൺֱͯ͠ҟৗݕΛߦ͏ -
24 ɾϚΠΫϩαʔϏεͰ ɺαʔϏεؒͷো ൖʹΑΓɺͲͷαʔ Ϗε͕ݪҼ͔ಛఆ͢Δ ͜ͱࠔ ɾؔ࿈ݚڀͯ͢ϝτ ϦοΫؒͷ૬ؔؔ ϝτϦοΫͱނোͷ૬ ؔؔΛΈ͍ͯΔ͕ɺ
ϩάॏཁ Root-Cause Metric Location for Microservice Systems via Log Anomaly Detection ɾDeeplog[8]ʹΑΓࢉग़ͨ͠ϩάͷҟৗείΞͱ ϝτϦοΫΛ߹ͤͯ͞ɺ૬ޓใྔʹج͍ͮͨ ૬ؔੳΛߦ͏ ɾඦͷࢦඪͷத͔Βฏۉ্ͯ͠Ґ15Ґ·Ͱʹࠜ ຊݪҼΛಛఆͰ͖ͨ -
26 ΦϯϥΠϯձͷ༷ࢠ ɾௌऺͷਓɺKeynoteͰ20ਓલޙɻฏۉ10ਓऑɺਓؾ͕ͳ͚Ε 1,2ਓͳͲɻඇৗʹগͳ͍ਓͰ͋Δͱײͨ͡ɻʢஶऀ؍ଌʣ ɾ࠲͕ෆࡏͷηογϣϯ͕͋ͬͨɻ࠲ͱൃදऀͷ3ਓ͏ͪ2ਓ͕͍ ͳ͍͜ͱ͋ͬͨɻ ɾUnderlineͷ৴ػೳͷ͍উखʹ͋ΓɻԻͱΧϝϥͷڞ༗ͷ On/Offঢ়ଶΛѲͰ͖ͣɺมߋ͢ΔʹҰ୴ୀग़͢Δඞཁ͕͋ͬͨɻ ɾ࠲Ҏ֎ͷௌऺʹΑΔ࣭ɺUnderlineͷQ&Aཝʹॻ͖ࠐ·ΕΔ͜ ͱ͕ଟ͔ͬͨɻ
ɾը৴࣌ʹ͏·͘Իͱը໘Λ৴Ͱ͖ͳ͍ɺԻׂ͕Ε͍ͯΔɺԻ ͕খ͍͞ɺԻ·ͨεϥΠυ৴్͕தͰ్ΕΔͳͲͷτϥϒϧ -
31 Early paper submission: March 1, 2021 Notification to early
papers: April 15, 2021 Normal paper submission due: May 5, 2021 Final notification to authors: June 17, 2021 Camera-ready paper due: July 1, 2021 དྷͷ։࠵༧ࠂ: IEEE SERVICES 2021 CHICAGO, ILLINOIS, USA September 5-10 ։࠵༧ఆͷࠃࡍձٞɺCLOUD/ICWS/SCC/SMDS ͷ4ͭ https://youtu.be/hwkGNGn4mt8 ΑΓҾ༻