performance isolation and fairness for multi tenant cloud
play

Performance Isolation and Fairness for Multi-Tenant Cloud Storage - PowerPoint PPT Presentation

Dec 21, 2023 •182 likes •1.2k views

PrincetonUniversity Performance Isolation and Fairness for Multi-Tenant Cloud Storage David Shue *, Michael Freedman*, and Anees Shaikh *Princeton IBM Research Setting: Shared Storage in the Cloud Y Y F F Z Z T T 2 Setting:

  1. Strawman: Queue Tenants By Single Resource Tenant A Tenant B VM VM VM VM VM VM PP Bandwidth limited Request Limited Controller WA W A2 W B2 RR RS FQ W A2 < W B2 out req out req 21

  2. Strawman: Queue Tenants By Single Resource Tenant A Tenant B VM VM VM VM VM VM PP Bandwidth limited Request Limited Controller WA W A2 W B2 bottleneck resource (out bytes) fair share RR RS FQ W A2 < W B2 out req out req 21

  3. Pisces: Queue Tenants By Dominant Resource Tenant A Tenant B VM VM VM VM VM VM PP Bandwidth limited Request Limited Controller WA W A2 W B2 RR RS FQ Track per-tenant W A2 < W B2 resource vector out req out req 22

  4. Pisces: Queue Tenants By Dominant Resource Tenant A Tenant B VM VM VM VM VM VM PP Bandwidth limited Request Limited Controller WA W A2 W B2 dominant resource fair share RR RS FQ Track per-tenant W A2 < W B2 resource vector out req out req 22

  5. Pisces: Queue Tenants By Dominant Resource Tenant A Tenant B VM VM VM VM VM VM PP Bandwidth limited Request Limited Controller WA W A2 W B2 dominant resource fair share RR RS FQ Track per-tenant W A2 < W B2 resource vector out req out req 22

  6. Pisces Mechanisms Solve For Global Fairness PP WA W A2 W B2 System Visibility RS d FQ Timescale 23

  7. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility RS d FQ minutes Timescale 23

  8. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility RS d FQ seconds minutes Timescale 23

  9. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility RR RS ... RR d SS local ... FQ SS microseconds seconds minutes Timescale 23

  10. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility RR Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  11. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility RR Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  12. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility RR Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  13. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  14. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  15. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  16. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  17. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  18. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  19. Pisces Mechanisms Solve For Global Fairness PP Controller global WA W A2 W B2 System Visibility fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  20. Pisces Mechanisms Solve For Global Fairness Maximum bottleneck PP flow weight exchange Controller global WA W A2 W B2 System Visibility fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  21. Pisces Mechanisms Solve For Global Fairness Maximum bottleneck PP flow weight exchange Controller global WA W A2 W B2 System Visibility FAST -TCP based replica selection fairness and capacity RR constraints Weight Allocations RS ... RR d SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  22. Pisces Mechanisms Solve For Global Fairness Maximum bottleneck PP flow weight exchange Controller global WA W A2 W B2 System Visibility FAST -TCP based replica selection fairness and capacity RR constraints Weight Allocations RS ... RR DRR token-based d DRFQ scheduler SS local ... Replica Selection FQ SS Policies microseconds seconds minutes Timescale 23

  23. Evaluation 24

  24. Evaluation • Does Pisces achieve (even) system-wide fairness? 24

  25. Evaluation • Does Pisces achieve (even) system-wide fairness? - Is each Pisces mechanism necessary for fairness? 24

  26. Evaluation • Does Pisces achieve (even) system-wide fairness? - Is each Pisces mechanism necessary for fairness? - What is the overhead of using Pisces? 24

  27. Evaluation • Does Pisces achieve (even) system-wide fairness? - Is each Pisces mechanism necessary for fairness? - What is the overhead of using Pisces? • Does Pisces handle mixed workloads? 24

  28. Evaluation • Does Pisces achieve (even) system-wide fairness? - Is each Pisces mechanism necessary for fairness? - What is the overhead of using Pisces? • Does Pisces handle mixed workloads? • Does Pisces provide weighted system-wide fairness? 24

  29. Evaluation • Does Pisces achieve (even) system-wide fairness? - Is each Pisces mechanism necessary for fairness? - What is the overhead of using Pisces? • Does Pisces handle mixed workloads? • Does Pisces provide weighted system-wide fairness? • Does Pisces provide local dominant resource fairness? 24

  30. Evaluation • Does Pisces achieve (even) system-wide fairness? - Is each Pisces mechanism necessary for fairness? - What is the overhead of using Pisces? • Does Pisces handle mixed workloads? • Does Pisces provide weighted system-wide fairness? • Does Pisces provide local dominant resource fairness? • Does Pisces handle dynamic demand? 24

  31. Evaluation • Does Pisces achieve (even) system-wide fairness? - Is each Pisces mechanism necessary for fairness? - What is the overhead of using Pisces? • Does Pisces handle mixed workloads? • Does Pisces provide weighted system-wide fairness? • Does Pisces provide local dominant resource fairness? • Does Pisces handle dynamic demand? • Does Pisces adapt to changes in object popularity? 24

  32. Evaluation • Does Pisces achieve (even) system-wide fairness? - Is each Pisces mechanism necessary for fairness? - What is the overhead of using Pisces? • Does Pisces handle mixed workloads? • Does Pisces provide weighted system-wide fairness? • Does Pisces provide local dominant resource fairness? • Does Pisces handle dynamic demand? • Does Pisces adapt to changes in object popularity? 25

  33. Pisces Achieves System-wide Per-tenant Fairness Ideal fair share: 110 kreq/s (1kB requests) Unmodified Membase 180 GET Requests (kreq/s) 160 140 120 100 80 60 40 0.57 MMR 20 0 10 20 30 40 50 60 70 80 Time (s) 8 Tenants - 8 Client - 8 Storage Nodes Zipfian object popularity distribution Min-Max Ratio: min rate/max rate (0,1]

  34. Pisces Achieves System-wide Per-tenant Fairness Ideal fair share: 110 kreq/s (1kB requests) Unmodified Membase Pisces 180 180 GET Requests (kreq/s) GET Requests (kreq/s) 160 160 140 140 120 120 100 100 80 80 60 60 40 40 0.57 MMR 0.98 MMR 20 20 0 0 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 Time (s) Time (s) 8 Tenants - 8 Client - 8 Storage Nodes Zipfian object popularity distribution Min-Max Ratio: min rate/max rate (0,1]

  35. Each Pisces Mechanism Contributes to System- wide Fairness and Isolation Unmodified Membase 180 GET Requests (kreq/s) 160 140 120 100 80 60 40 0.57 MMR 20 0 10 20 30 40 50 60 70 80 Time (s) 27

  36. Each Pisces Mechanism Contributes to System- wide Fairness and Isolation Unmodified Membase 180 GET Requests (kreq/s) 160 140 120 100 80 60 40 0.57 MMR 20 0 10 20 30 40 50 60 70 80 Time (s) 2x vs 1x demand 180 GET Requests (kreq/s) 160 140 120 100 80 60 40 0.36 MMR 20 0 10 20 30 40 50 60 70 80 Time (s) 27

  37. Each Pisces Mechanism Contributes to System- wide Fairness and Isolation Unmodified Membase FQ 180 180 GET Requests (kreq/s) 160 160 140 140 120 120 100 100 80 80 60 60 40 40 0.57 MMR 0.59 MMR 20 20 0 0 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 90 Time (s) 2x vs 1x demand 180 180 GET Requests (kreq/s) 160 160 140 140 120 120 100 100 80 80 60 60 40 40 0.58 MMR 0.36 MMR 20 20 0 0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 Time (s) 27

  38. Each Pisces Mechanism Contributes to System- wide Fairness and Isolation Unmodified Membase FQ FQ PP 180 180 180 GET Requests (kreq/s) 160 160 160 140 140 140 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 0.57 MMR 0.59 MMR 0.64 MMR 20 20 20 0 0 0 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 Time (s) 2x vs 1x demand 180 180 180 GET Requests (kreq/s) 160 160 160 140 140 140 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 0.58 MMR 0.74 MMR 0.36 MMR 20 20 20 0 0 0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 Time (s) 27

  39. Each Pisces Mechanism Contributes to System- wide Fairness and Isolation Unmodified Membase FQ FQ PP WA 180 180 180 GET Requests (kreq/s) 160 160 160 140 140 140 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 0.57 MMR 0.59 MMR 0.64 MMR 0.93 MMR 20 20 20 0 0 0 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 Time (s) 2x vs 1x demand 180 180 180 GET Requests (kreq/s) 160 160 160 140 140 140 120 120 120 100 100 100 80 80 80 60 60 60 40 40 40 0.58 MMR 0.74 MMR 0.96 MMR 0.36 MMR 20 20 20 0 0 0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 Time (s) 27

  40. Each Pisces Mechanism Contributes to System- wide Fairness and Isolation Unmodified Membase FQ RS FQ PP FQ PP WA 180 180 180 180 GET Requests (kreq/s) 160 160 160 160 140 140 140 140 120 120 120 120 100 100 100 100 80 80 80 80 60 60 60 60 40 40 40 40 0.57 MMR 0.59 MMR 0.64 MMR 0.93 MMR 0.90 MMR 20 20 20 20 0 0 0 0 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 Time (s) 2x vs 1x demand 180 180 180 180 GET Requests (kreq/s) 160 160 160 160 140 140 140 140 120 120 120 120 100 100 100 100 80 80 80 80 60 60 60 60 40 40 40 40 0.58 MMR 0.74 MMR 0.96 MMR 0.89 MMR 0.36 MMR 20 20 20 20 0 0 0 0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 Time (s) 27

  41. Each Pisces Mechanism Contributes to System- wide Fairness and Isolation Unmodified Membase FQ RS FQ PP FQ PP WA WA 180 180 180 180 GET Requests (kreq/s) 160 160 160 160 140 140 140 140 120 120 120 120 100 100 100 100 80 80 80 80 60 60 60 60 40 40 40 40 0.57 MMR 0.59 MMR 0.64 MMR 0.93 MMR 0.90 MMR 0.98 MMR 20 20 20 20 0 0 0 0 10 20 30 40 50 60 70 80 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 Time (s) 2x vs 1x demand 180 180 180 180 GET Requests (kreq/s) 160 160 160 160 140 140 140 140 120 120 120 120 100 100 100 100 80 80 80 80 60 60 60 60 40 40 40 40 0.58 MMR 0.74 MMR 0.96 MMR 0.89 MMR 0.97 MMR 0.36 MMR 20 20 20 20 0 0 0 0 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 90 10 20 30 40 50 60 70 80 Time (s) 27

  42. Pisces Imposes Low-overhead Aggregate System Throughput 3500 > 19% GET Requests (kreq/s) 2625 1750 < 5% 875 0 1kB Requests 10B Requests Unmodified Membase Pisces 28

  43. Pisces Achieves System-wide Weighted Fairness 4 heavy hitters 20 moderate demand 40 low demand 29

  44. Pisces Achieves System-wide Weighted Fairness 4 heavy hitters 20 moderate demand 40 low demand 160 GET Requests (kreq/s) 140 120 100 80 100x weight (4) 60 10x weight (20) 40 1x weight (40) 20 0 25 30 35 40 45 50 55 60 Time (s) 29

  45. Pisces Achieves System-wide Weighted Fairness 0.98 MMR 0.89 MMR 0.91 MMR 4 heavy hitters 20 moderate demand 40 low demand 160 GET Requests (kreq/s) 140 120 100 80 100x weight (4) 60 10x weight (20) 40 1x weight (40) 20 0 25 30 35 40 45 50 55 60 Time (s) 29

  46. Pisces Achieves System-wide Weighted Fairness 0.98 MMR 0.89 MMR 0.91 MMR 4 heavy hitters 20 moderate demand 40 low demand 160 GET Requests (kreq/s) 140 120 100 0.91 MMR 80 100x weight (4) 60 10x weight (20) 40 1x weight (40) 20 0 25 30 35 40 45 50 55 60 Time (s) 29

  47. Pisces Achieves System-wide Weighted Fairness 0.98 MMR 0.89 MMR 0.91 MMR 4 heavy hitters 20 moderate demand 40 low demand 160 GET Requests (kreq/s) 140 120 100 0.91 MMR 80 100x weight (4) 60 10x weight (20) 40 1x weight (40) 20 0.56 MMR 0 25 30 35 40 45 50 55 60 Time (s) 29

  48. Pisces Achieves Dominant Resource Fairness 1kB workload 10B workload bandwidth limited request limited 30

  49. Pisces Achieves Dominant Resource Fairness 1kB workload 10B workload bandwidth limited request limited 800 700 Bandwidth (Mb/s) 600 500 400 300 200 1kB bandwidth limited 10B request limited 100 0 20 25 30 35 40 45 50 55 60 Time (s) 30

  50. Pisces Achieves Dominant Resource Fairness 1kB workload 10B workload bandwidth limited request limited GET Requests (kreq/s) 800 300 700 Bandwidth (Mb/s) 250 600 200 500 400 150 300 100 200 1kB bandwidth limited 50 10B request limited 100 0 0 20 25 30 35 40 45 50 55 60 20 25 30 35 40 45 50 55 60 Time (s) Time (s) 30

  51. Pisces Achieves Dominant Resource Fairness 1kB workload 10B workload bandwidth limited request limited GET Requests (kreq/s) 800 76% of bandwidth 76% of request rate 300 700 Bandwidth (Mb/s) 250 600 200 500 400 150 300 100 200 1kB bandwidth limited 50 10B request limited 100 0 0 20 25 30 35 40 45 50 55 60 20 25 30 35 40 45 50 55 60 Time (s) Time (s) 30

  52. Pisces Achieves Dominant Resource Fairness 1kB workload 10B workload bandwidth limited request limited GET Requests (kreq/s) 800 76% of bandwidth 76% of request rate 300 700 Bandwidth (Mb/s) 250 600 200 500 400 150 24% of request rate 300 100 200 1kB bandwidth limited 50 10B request limited 100 0 0 20 25 30 35 40 45 50 55 60 20 25 30 35 40 45 50 55 60 Time (s) Time (s) 30

  53. Pisces Adapts to Dynamic Demand Tenant Demand Constant Diurnal (2x wt) Bursty 31

  54. Pisces Adapts to Dynamic Demand Tenant Demand Constant Diurnal (2x wt) Bursty 200 GET Requests (kreq/s) 150 100 50 0 0 10 20 30 40 50 60 70 80 90 Time (s) 31

  55. Pisces Adapts to Dynamic Demand Tenant Demand Constant Diurnal (2x wt) Bursty 200 GET Requests (kreq/s) 150 ~2x 100 50 0 0 10 20 30 40 50 60 70 80 90 Time (s) 31

  56. Pisces Adapts to Dynamic Demand Tenant Demand Constant Diurnal (2x wt) Bursty 200 GET Requests (kreq/s) even 150 ~2x 100 50 0 0 10 20 30 40 50 60 70 80 90 Time (s) 31

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Fine-Grained Isolation for Scalable, Dynamic, Multi-tenant Edge Clouds Yuxin Ren, Guyue Liu,

Fine-Grained Isolation for Scalable, Dynamic, Multi-tenant Edge Clouds Yuxin Ren, Guyue Liu,

Fine-Grained Isolation for Scalable, Dynamic, Multi-tenant Edge Clouds Yuxin Ren, Guyue Liu, Vlad Nitu, Wenyuan Shao, Riley Kennedy, Gabriel Parmer, Timothy Wood, Alain Tchana Presented by: Vlad Nitu Edge Cloud 40 million IoT devices in

266 views • 26 slides
Large objects in the Cloud Thursday, 11 April 13 Riak Cloud Storage Cloud Storage software

Large objects in the Cloud Thursday, 11 April 13 Riak Cloud Storage Cloud Storage software

Large objects in the Cloud Thursday, 11 April 13 Riak Cloud Storage Cloud Storage software backed by Riak Simple API Multi-tenant, Per-tenant Reporting Pluggable Authentication Multi Data Center Replication (Enterprise)

369 views • 18 slides
Distributed multi-tenant cloud/fog and heterogeneous SDN/NFV orchestration for 5G services Ricard

Distributed multi-tenant cloud/fog and heterogeneous SDN/NFV orchestration for 5G services Ricard

Distributed multi-tenant cloud/fog and heterogeneous SDN/NFV orchestration for 5G services Ricard Vilalta, A. Mayoral, Raul Muoz, Ramon Casellas, Ricardo Martnez The need for generic control functions and a Transport API The NBI of the

679 views • 15 slides
Feasibility of ILA as Network Virtualisation Overlay in multi-domain, multi-tenant Cloud Tako

Feasibility of ILA as Network Virtualisation Overlay in multi-domain, multi-tenant Cloud Tako

Feasibility of ILA as Network Virtualisation Overlay in multi-domain, multi-tenant Cloud Tako Marks Bsc. 04-07-2017 System and Network Engineering University of Amsterdam 1 Introduction A Network Virtualisation Overlay (NVO) is a collection

614 views • 29 slides
Seawall: Performance Isolation for Cloud Datacenter Networks Alan Shieh Cornell University

Seawall: Performance Isolation for Cloud Datacenter Networks Alan Shieh Cornell University

Seawall: Performance Isolation for Cloud Datacenter Networks Alan Shieh Cornell University Srikanth Kandula Albert Greenberg Changhoon Kim Microsoft Research Cloud datacenters: Benefits and obstacles Moving to the cloud has manageability,

221 views • 19 slides
FIT: A Distributed Database Performance Tradeoff Jose M. Faleiro, Yale University Daniel J.

FIT: A Distributed Database Performance Tradeoff Jose M. Faleiro, Yale University Daniel J.

FIT: A Distributed Database Performance Tradeoff Jose M. Faleiro, Yale University Daniel J. Abadi, Yale University Presented by Bojun Wang FAIRNESS THROUGHPUT ISOLATION 1 Isolation v.s. Throughput and Fairness Strong isolation

542 views • 20 slides
Institute for Cyber Security Multi-Tenant Access Control for Cloud Services PhD Dissertation

Institute for Cyber Security Multi-Tenant Access Control for Cloud Services PhD Dissertation

Institute for Cyber Security Multi-Tenant Access Control for Cloud Services PhD Dissertation Defense Bo Tang Committee Members: Dr. Ravi Sandhu, Chair Dr. Kay Robbins Dr. Gregory White Dr. Weining Zhang Dr. Jaehong Park 07/31/2014

539 views • 51 slides
Software-Defined Networking Architecture Framework for Multi-Tenant Enterprise Cloud Environments

Software-Defined Networking Architecture Framework for Multi-Tenant Enterprise Cloud Environments

Introduction Solutions Evaluation Summary Software-Defined Networking Architecture Framework for Multi-Tenant Enterprise Cloud Environments Aryan TaheriMonfared Department of Electrical Engineering and Computer Science University of

1.51k views • 92 slides
Keystone and SAML in multi-tenant environments Two concepts get together Alex Alex Stel

Keystone and SAML in multi-tenant environments Two concepts get together Alex Alex Stel

Keystone and SAML in multi-tenant environments Two concepts get together Alex Alex Stel ellwag, Open Telekom Cloud Architect ,T-Systems Yue uefen eng Pan an, Senior Cloud Solutions Architect, Huawei Open Telekom Cloud 26.06.2017 1

294 views • 28 slides
X-Containers: Breaking Down Barriers to Improve Performance and Isolation of Cloud-Native

X-Containers: Breaking Down Barriers to Improve Performance and Isolation of Cloud-Native

X-Containers: Breaking Down Barriers to Improve Performance and Isolation of Cloud-Native Containers Zhiming Shen Cornell University Joint work with Zhen Sun, Gur-Eyal Sela, Eugene Bagdasaryan, Christina Delimitrou, Robbert Van Renesse, Hakim

266 views • 26 slides
Offloading data plane functions to the multi-tenant cloud infrastructure using P4 Tomasz Osiski

Offloading data plane functions to the multi-tenant cloud infrastructure using P4 Tomasz Osiski

Offloading data plane functions to the multi-tenant cloud infrastructure using P4 Tomasz Osiski / Orange, WUT Mateusz Kossakowski / Orange, WUT Halina Tarasiuk / WUT Roland Picard / Orange 1 Orange Confidential

529 views • 16 slides
Easy multi-tenant Kubernetes RWX storage with Cloud Provider OpenStack and Manila CSI Tom Barron

Easy multi-tenant Kubernetes RWX storage with Cloud Provider OpenStack and Manila CSI Tom Barron

Easy multi-tenant Kubernetes RWX storage with Cloud Provider OpenStack and Manila CSI Tom Barron tbarron@redhat.com Victoria Martinez de la Cruz victoria@redhat.com WHAT ARE WE GOING TO SEE TODAY Game plan What is Manila CSI? Why RWX

1.36k views • 21 slides
2DFQ: Two-Dimensional Fair Queuing for Multi-Tenant Cloud Services Jona nathan an Ma Mace

2DFQ: Two-Dimensional Fair Queuing for Multi-Tenant Cloud Services Jona nathan an Ma Mace

2DFQ: Two-Dimensional Fair Queuing for Multi-Tenant Cloud Services Jona nathan an Ma Mace Peter Bodik Madanlal Musuvathi Brown University Microsoft Microsoft Rodrigo Fonseca Krishnan Varadarajan Brown University Microsoft Containers

1.93k views • 179 slides
Resource Management for Isolation Enhanced Cloud Services Enhanced Cloud Services Himanshu Raj ,

Resource Management for Isolation Enhanced Cloud Services Enhanced Cloud Services Himanshu Raj ,

Resource Management for Isolation Enhanced Cloud Services Enhanced Cloud Services Himanshu Raj , Ripal Nathuji, Abhishek Singh, Paul England XCG, Microsoft Research Motivation Isolation issues in shared-resource computing infrastructures

408 views • 13 slides
6.888 Lecture 6: Network Performance Isola8on Mohammad Alizadeh Spring 2016 1 Mul8-tenant

6.888 Lecture 6: Network Performance Isola8on Mohammad Alizadeh Spring 2016 1 Mul8-tenant

6.888 Lecture 6: Network Performance Isola8on Mohammad Alizadeh Spring 2016 1 Mul8-tenant Cloud Data Centers Shared infrastructure between mul8ple tenants/apps Lack of Performance Predictability GAE memcache read 100 values Unpredictable

730 views • 31 slides
Gatekeeper: Supporting Bandwidth Guarantees for Multi-tenant Datacenter Networks Henrique

Gatekeeper: Supporting Bandwidth Guarantees for Multi-tenant Datacenter Networks Henrique

Gatekeeper: Supporting Bandwidth Guarantees for Multi-tenant Datacenter Networks Henrique Rodrigues , Yoshio Turner , Jose Renato Santos , Paolo Victor , Dorgival Guedes HP Labs WIOV 2011, Portland, OR The Problem: Network Performance

846 views • 37 slides
Ubiquitous and Mobile Computing CS 525M: Mobile MapReduce: Minimizing Response Time of Computing

Ubiquitous and Mobile Computing CS 525M: Mobile MapReduce: Minimizing Response Time of Computing

Ubiquitous and Mobile Computing CS 525M: Mobile MapReduce: Minimizing Response Time of Computing Intensive Mobile Applications Vijay Sukhadeve Computer Science Dept. Worcester Polytechnic Institute (WPI) Introduction/motivation overall users

548 views • 18 slides
Learnin ning g Maximal al Marginal nal Re Releva vanc nce e Model via Directly ctly

Learnin ning g Maximal al Marginal nal Re Releva vanc nce e Model via Directly ctly

Learnin ning g Maximal al Marginal nal Re Releva vanc nce e Model via Directly ctly Optim imizi izing ng Diver ersi sity ty Ev Evalua uatio tion Measur ures es Long Xia , Jun Xu, Yanyan Lan, Jiafeng Guo, Xueqi Cheng Key

882 views • 30 slides
Axiomatic Foundations of Multiplier Preferences Tomasz Strzalecki Multiplier preferences

Axiomatic Foundations of Multiplier Preferences Tomasz Strzalecki Multiplier preferences

Axiomatic Foundations of Multiplier Preferences Tomasz Strzalecki Multiplier preferences Expected Utility inconsistent with observed behavior We (economists) may not want to fully trust any probabilistic model. Hansen and Sargent:

879 views • 84 slides
CSE 311: Foundations of Computing I Lecture 1: Propositional Logic Some Perspective Computer

CSE 311: Foundations of Computing I Lecture 1: Propositional Logic Some Perspective Computer

CSE 311: Foundations of Computing I Lecture 1: Propositional Logic Some Perspective Computer Science and Engineering Theory Programming CSE 14x CSE 311 Hardware About the Course We will study the theory needed for CSE: Logic : How can we

454 views • 44 slides
Modelling imperfect maintenance and repair of components under competing risk Helge Langseth and

Modelling imperfect maintenance and repair of components under competing risk Helge Langseth and

Modelling imperfect maintenance and repair of components under competing risk Helge Langseth and Bo H. Lindqvist { helgel, bo } @math.ntnu.no. Department of Mathematical Sciences, Norwegian University of Science and Technology, Trondheim,

544 views • 18 slides
Prismatic HTGR Dominique Hittner www.nc2i.eu NC2I is one of SNETPs strategic technological

Prismatic HTGR Dominique Hittner www.nc2i.eu NC2I is one of SNETPs strategic technological

Prismatic HTGR Dominique Hittner www.nc2i.eu NC2I is one of SNETPs strategic technological pillars, mandated to coordinate the demonstration of high temperature nuclear cogeneration. www.snetp.eu Contents Core Internals Reactor

843 views • 41 slides
FARM: A Prototyping Environment for Tightly-Coupled, Heterogeneous Architectures Tayo Oguntebi,

FARM: A Prototyping Environment for Tightly-Coupled, Heterogeneous Architectures Tayo Oguntebi,

FARM: A Prototyping Environment for Tightly-Coupled, Heterogeneous Architectures Tayo Oguntebi, Sungpack Hong, Jared Casper, Nathan Bronson Christos Kozyrakis, Kunle Olukotun Outline Motivation The Stanford FARM Using FARM

755 views • 60 slides
Water Vapor Seasonal Variability: AIRS L3 results vs. climate models Thomas Hearty, George

Water Vapor Seasonal Variability: AIRS L3 results vs. climate models Thomas Hearty, George

National Aeronautics and Space Administration Jet Propulsion Laboratory California Institute of Technology Pasadena, California Water Vapor Seasonal Variability: AIRS L3 results vs. climate models Thomas Hearty, George Aumann, Duane Waliser,

482 views • 20 slides

More recommend