Optimizing Service Selection and Load Balancing in Multi-Cluster Microservice Systems with MCOSS
This research delves into optimizing service selection and load balancing in multi-cluster microservice systems, tackling challenges and proposing solutions through collaboration with industry experts. Explore the intricacies of monolithic vs. micro-service applications, Kubernetes container orchestration, and multi-cluster deployment complexities.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.
E N D
Presentation Transcript
Optimizing Service Selection and Load Balancing in Multi-Cluster Microservice Systems with MCOSS Daniel Bachar Reichman University, Israel Joint work with Anat Bremler-Barr (Tel Aviv U.) and David Hay (The Hebrew U.). This research is in collaboration with Red Hat.
Outline Background The multi-cluster service selection problem Solution - Model and Architecture Evaluation Conclusions 2
Monolithic Application Product Logic Review Logic Details Logic Rating Logic Requests 3
Micro-service Application Requests Requests Requests Product Review Rating Requests Details 4
Kubernetes Container Orchestration Infrastructure Abstraction Pod Container Container Container 5
Kubernetes Container Orchestration Infrastructure Abstraction Pod Container Container Container 6
Kubernetes Container Orchestration Infrastructure Abstraction Node - Compute Pod No Pod 7
Kubernetes Container Orchestration Infrastructure Abstraction Service Pod Pod Pod 8
Kubernetes Container Orchestration Infrastructure Abstraction Service 9
Kubernetes - Building Blocks Kubernetes Cluster Product Review Rating Front End Front End Other Applications Details 10
Outline Background The multi-cluster service selection problem Solution - Model and Architecture Evaluation Conclusions 11
Multi Cluster Deployment Kubernetes Cluster Kubernetes Cluster Kubernetes Cluster Kubernetes Cluster Given a request from a source service, which target service Multi-cluster deployment introduces traffic pricing and performance challenges Product Product should be selected? Review Review 12
The Multi-Cluster Selection Problem Objective Deployment Constraints 13
Architecture: Data Plane WRR Cluster 2 Cluster 1 DNS DNS (2) (1) Rating Review Product (3) 14
Architecture: Control Plane Objective Broker Constraints Cluster 1 Cluster 2 DNS DNS Metrics Metrics Product Review Rating 15
The Multi-Cluster Selection Problem Objective Constraints 16
Model: Constraints One Hop Optimization Cluster 1) Demand: Requests cannot be dropped Product Review Product Review 2) Capacity: Service cannot handle more than its capacity Product Review 17
Model: Objective One Hop Optimization W1 = 0.3 Product Review W2 = 0.7 Objective: Minimize the cost of the requests W3 = 1 Product Review W4 = 0.6 Product Review W5 = 0.4 18
Cost functions - ILP Mixing Latency and Billing cost Price per GB in USD$ Latency between clusters (ms) 0 142 142 0 0.08 0.15 142 0 3 0.08 0 0.15 142 3 0 0.11 0.11 0 19 19
Cost functions - QP Considering response time QP, Proof see paper 20
Outline Background The multi-cluster service selection problem Solution - Model and Architecture Evaluation Conclusions 21
Sample App Price per GB in USD$ 0 0.08 0.15 0.08 0 0.15 Product Review Rating 0.11 0.11 0 Latency between clusters (ms) Details 0 142 142 142 0 3 142 3 0 22
Evaluation: Performance - Local decision at each cluster. - Prune to herd behavior. 23
Outline Background The multi-cluster service selection problem Solution - Model and Architecture Evaluation Conclusions 25
Summary Service selection problem Challenges 26
Multi-hop / Full service chain Product Review Rating Details 28
Scalability Edge (5G, etc) Our case 29
Thank you! bachar.daniel@post.runi.ac.il Simulator Implementation Deepness lab 30
