Ninja Meeting 2/15/2000 Sam Madden

Slides:



Advertisements
Similar presentations
Transport Layer3-1 Transport Overview and UDP. Transport Layer3-2 Goals r Understand transport services m Multiplexing and Demultiplexing m Reliable data.
Advertisements

CPSC 441: Intro, UDP1 Transport Layer Instructor: Carey Williamson Office: ICT Class Location:
Point-to-Point Network Switching Advanced Computer Networks.
1 8. Safe Query Languages Safe program – its semantics can be at least partially computed on any valid database input. Safety is tied to program verification,
CPSC 441: Intro, UDP1 Instructor: Anirban Mahanti Office: ICT Class Location: ICT 121 Lectures: MWF 12:00 – 12:50 Notes.
Point-to-Point Network Switching Computer Networks Computer Networks Spring 2012 Spring 2012.
Deliver Multimedia Streams with Flexible QoS via a Multicast DAG Yu Cai 02/26/2004.
Internetworking Fundamentals (Lecture #2) Andres Rengifo Copyright 2008.
IP-UDP-RTP Computer Networking (In Chap 3, 4, 7) 건국대학교 인터넷미디어공학부 임 창 훈.
SERIAL BUS COMMUNICATION PROTOCOLS
SpaceWire-RT Steve Parkes, Albert Ferrer-Florit
OPERATOR PRECEDENCE PARSING
Mukesh N. Tekwani Elphinstone College Mumbai
Marwan Al-Namari Week 4. Physical Path Application Presentation Session Transport Network Data Link Physical Application Presentation Session Transport.
Introduction to Networks CS587x Lecture 1 Department of Computer Science Iowa State University.
Wireless Access and Terminal Mobility in CORBA Dimple Kaul, Arundhati Kogekar, Stoyan Paunov.
Multi-part Messages in KMIP John Leiseboer, QuintessenceLabs.
Switching Techniques Dr. Sanjay P. Ahuja, Ph.D. Fidelity National Financial Distinguished Professor of CIS School of Computing, UNF.
Chapter 15 – Part 2 Networks The Internal Operating System The Architecture of Computer Hardware and Systems Software: An Information Technology Approach.
ECEN “Internet Protocols and Modeling”, Spring 2012 Course Materials: Papers, Reference Texts: Bertsekas/Gallager, Stuber, Stallings, etc Class.
Chapter 3: Transport Layer Our goals: r understand principles behind transport layer services: m multiplexing/demultipl exing m reliable data transfer.
Transport Layer3-1 Chapter 3 Transport Layer Computer Networking: A Top Down Approach Featuring the Internet, 3 rd edition. Jim Kurose, Keith Ross Addison-Wesley,
Department of Electronic Engineering City University of Hong Kong EE3900 Computer Networks Protocols and Architecture Slide 1 Use of Standard Protocols.
MULTIPLEXING/DEMULTIPLEXING, CONNECTIONLESS TRANSPORT.
ATM Protocol Architecture Mode. ATM architecture uses a logical reference model to describe its functions ATM functions correspond to physical layer and.
COMPUTER NETWORKS Hwajung Lee. Image Source:
Distributed Computing & Embedded Systems Chapter 4: Remote Method Invocation Dr. Umair Ali Khan.
Chapter 3 Transport Layer
Transport Layer Slides are originally from instructor: Carey Williamson at University of Calgary Very minor modification are made Notes derived from “Computer.
S. Sudarshan CS632 Course, Mar 2004 IIT Bombay
Chapter 3 outline 3.1 Transport-layer services
Object-oriented and Structured System Models
Point-to-Point Network Switching
03 – Remote invoaction Request-reply RPC RMI Coulouris 5
EE 122: Lecture 19 (Asynchronous Transfer Mode - ATM)
Network Architecture Layered system with alternative abstractions available at a given layer.
Distributed database approach,
UNIT-V Transport Layer protocols for Ad Hoc Wireless Networks
Basics of Intrusion Detection
NSI Topology Thoughts on how topology fits into the NSI architecture
Telemedicine.
Introduction and Overview of Network and Telecommunications (contd.)
Data Networking via LANs Packets and Frames
Software Design and Architecture
Wireless Sensor Network Architectures
Understand the OSI Model Part 2
Interquery Parallelism
CS 1652 Jack Lange University of Pittsburgh
Net431:advanced net services
Transport Layer Our goals:
September 19th, 2013 CS1652 Jack Lange University of Pittsburgh
Streaming Sensor Data Fjord / Sensor Proxy Multiquery Eddy
An aggregation mechanism
IV. Convolutional Codes
CSE S. Tanimoto Paradigms
Chapter 2. Protocols and Architecture
Design Yaodong Bi.
Chapter 5 Transport Layer Introduction
OSI Reference Model Unit II
Computer Networks Protocols
Chapter 3 Transport Layer
OSI Model 7 Layers 7. Application Layer 6. Presentation Layer
Last Class: Communication in Distributed Systems
Architectural Mismatch: Why reuse is so hard?
Point-to-Point Network Switching
OPERATOR GRAMMAR No Ɛ-transition. No two adjacent non-terminals. Eg.
Transport Layer 9/22/2019.
EEL 5718 Computer Communications
Transport Layer Our goals:
Presentation transcript:

Ninja Meeting 2/15/2000 Sam Madden Messages in Paths Ninja Meeting 2/15/2000 Sam Madden

Telegraph Extensions Message (Tuple) Based Transmission Layer

Telegraph Extensions Message (Tuple) Based Transmission Layer Aggregators/Broadcasters

Telegraph Extensions Message (Tuple) Based Transmission Layer Aggregators/Broadcasters Operator Typing System Interoperability Constraints

Telegraph Extensions Message (Tuple) Based Transmission Layer Aggregators/Broadcasters Operator Typing System Interoperability Constraints Mechanisms for Performance Evaluation

Telegraph Extensions Message (Tuple) Based Transmission Layer Aggregators/Broadcasters Operator Typing System Interoperability Constraints Mechanisms for Performance Evaluation

Messages vs. Streams Streams Messages Partitioned Data Unpartitioned Data

Messages vs. Streams Streams Messages Partitioned Data Push or Pull Data Unpartitioned Data Push out data

Messages vs. Streams Streams Messages Partitioned Data Push or Pull Data Higher Overhead Unpartitioned Data Push out data

Messages vs. Streams Streams Messages Partitioned Data Push or Pull Data Higher Overhead More Complicated Programming Model Unpartitioned Data Push out data

Messages vs. Streams Streams Messages Partitioned Data Push or Pull Data Higher Overhead More Complicated Programming Model Unpartitioned Data Push out data No: Type Checking Admission Control Aggregation End-to-End Guarantees

Benefits of Messages Aggregation Improved End-To-End Guarantees Structural Type Checking (not Semantics) Admission Control Pull-based consumption Natural model for databases, Web

Architecture Messages are tuples: <Idenitifer id, int Length, Byte[] data> Identifier may encapsulate semantic information, sequence number, etc.

Architecture Messages are tuples: <Idenitifer id, int Length, Byte[] data> Identifier may encapsulate semantic information, sequence number, etc. Operators Specify Multiple Inputs {<Identifier id, Class C>, … } Incoming messages are packaged as type C via (de)serialization methods.

Architecture #2 Class C may be unstructured (e.g. String) Low overhead Structured classes allow Data verification Admission Control

Architecture #3 Operators can request tuples from an input or push tuples down an output. Generates events on predceding/following node

Architecture #3 Operators can request tuples from an input or push tuples down an output. Generates events on predceding/following node Events model implies VSpace

Architecture #3 Operators can request tuples from an input or push tuples down an output. Generates events on predceding/following node Events model implies VSpace Maybe?

On Top Of Streams Streams (not RMI) are underlying transmission protocol

On Top Of Streams Streams (not RMI) are underlying transmission protocol Some operators still “stream-based”

On Top Of Streams Streams (not RMI) are underlying transmission protocol Some operators still “stream-based” Streams provide Low-level error detection Control over threads/operator Guaranteed reasonable performance