Exploring Emerging Technologies in Cloud Computing
Cloud computing revolutionizes accessibility to computing resources through shared pools. Next-generation applications, edge computing, and web applications on the cloud are shaping the future. CloudPath offers a platform for executing third-party applications across distributed data centers. Real-world examples demonstrate the versatility of cloud-based solutions.
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. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
E N D
Presentation Transcript
Seyed Hossein Mortazavi mortazavi@cs.toronto.edu
Cloud Computing Enables ubiquitous, and convenient access to a shared pool of configurable computing resources Servers Storage Applications Services Network 2
Next Generation Apps Next generation applications require lower latencies, better response time Virtual/Augmented Reality, wearable devices, safety-critical applications, etc. They need high bandwidth Video analysis, Internet of Things(IoT) devices Other requirements: privacy, exact location detection, scalability, 3
Edge Computing Bring the cloud closer to the users (Mobile Edge Computing, Cloudlets, Fog Computing) Computation and Storage units are placed near the user On the Edge of the network
Web Applications on the cloud Collection of independent stateless handlers REST (flexibility) anywhere Scalability 10ms Fault tolerance 100 ms 20 ms What Enables this model? Shared storage layer 5
CloudPath A platform that enables the execution of 3rd party applications on a progression of datacenters deployed along the geographical span of the network Separation between application code and data Developers: organize applications as a collection of stateless functions CloudPath: on-demand replication of code and data Provides a common runtime on all cloud nodes 6
CloudPath Each node consists of: Common execution environment --> PathExecute On-Demand deployment --> PathDeploy Routing --> PathRoute Common storage layer --> PathStore Monitoring --> PathMonitor 7
Example Cloud App1 App2 F1: Record flight info F1 F2 F3 F2 F1 Core 1 Core 2 F2: Get sensory data, make decisions on flight path App1 App2 F3 F2 F1 F2 Edge 3 Edge 1 Edge 2 F3: Get frame, do object avoidance App1 App1 F3 F3 8
Pathstore: A Distributed Storage Layer For The Edge PathStore is a data storage layer for a multi-tier cloud architecture Hierarchy of data stores Each PathStore node runs an independent Cassandra ring and PathStore copies data between rings Support Eventual consistency Extension called SessionStore supports session consistency 9
PathStore Design Data store at the root of the hierarchy is persistent Cloud Other act as temporary partial replicas Data replicated on-demand Core 1 Core 2 All reads and writes executed against local replica Edge 1 Edge 2 Edge 3 CQL based interface Row-level replication
Example Cloud Marbles Table select * from marbles where color = blue select * from marbles Core 1 Core 2 select * from marbles Marbles Table Marbles Table select * from marbles where color = blue select * from marbles where color = green Edge 1 Edge 2 Edge 3 Marbles Table Marbles Table Marbles Table select * from marbles select * from marbles where color = green select * from marbles where color = blue 11
SessionStore SessionStore ensures session consistency on a top of otherwise eventually consistent replicas. Enforces session consistency by grouping related data accesses into a session, and using a session-aware reconciliation algorithm to reconcile only the data that is relevant to the session when switching between data centers. Only replicates relevant data (row level) Only replicates to the relevant destination 12
SessionStore Cloud Transfer only Only move Switching between Move all data for the application session data newer session data neighbors Stamps Table Marbels Table Core 1 Core 2 Stamps Table Marbels Table Stamps Table Marbels Table Edge 1 Edge 2 Edge 3 Stamps Table Marbels Table Stamps Table Marbels Table Marbels Table Stamps Table select * from marbles where color=blue select * from marbles Insert green_marble into marbles 14
Session Tracking We identify each session using a Session Token, or stoken. Stoken consists of a four fields: A unique session id (SID) timestamp current replica status The stokenis encrypted and signed to prevent forging and misrepresentation. To keep track of data related to a session, a command-Cache is added to each PathStore replica that stores all the CQL SELECT statements that were run on behalf of a session 15
Network Requirements Routing requests for applications based on functions Synchronizing clocks between Datacenters Providing locking service fordata Better guarantees for data 16
Questions Contact: mortazavi@cs.toronto.edu 17
