What we do in Grids e Science CyberInfrastructure and Peer to Peer Networks

What we do i

Grids e-Scienc

CyberInfrastructure

and Peer-to-Pee

Network

Los Alamo

September 23 2003

Geoffrey Fox

Community Grids Lab

Indiana University

What do we do?

•

Portals

: co-chair of Grid Computing Environment

research group at GGF

•

Metadata

: co-chair at GGF of Semantic Grid RG

– Apply to Earth Science using GML (Geography mark up)

•

Collaboration

: built a Web Service based collaboration

environment sharing applications and audio/video

conferencing to desktops and

PDA

’s

•

Web Service

model for all applications

•

Messaging

: Open “Grid Messaging System”

NaradaBrokering linking P2P and (Cellular) Grid

•

Autonomic services

using managed messages

•

Applications

– CrisisGrid – Indiana and openGIS Consortium

– SERVOGrid for Earthquake Science

Collage of Portals

Earthquakes – NAS Fusion – DoE

“GridMPI” v. NaradaBrokering

n In parallel computing, MPI and PVM provided “all the features

one needed’ for inter-node messaging

n NB aims to play same role for the Grid but the requirements and

constraints are very different

• NB is not MPI ported to a Grid/Globus environment

n Typically MPI aiming at microsecond latency but for Grid, time

scales are different

• 100 millisecond quite normal network latency

• 30 millisecond typical packet time sensitivity (this is one audio or video

frame) but even here can buffer 10-100 frames on client (conferencing to streaming)

• 1 millisecond is time for a Java server to “think”

n Jitter in latency (transit time through broker) due to routing,

processing (in NB) or packet loss recovery is important property

n Grids need and can use software supported message functions and

trade-offs between hardware and software routing different from

NaradaBrokering

n

Based on a network of cooperating

broker nodes

• Cluster based architecture allows system to scale in size

n

Originally designed to provide

uniform software

multicast

to support real-time collaboration linked to

publish-subscribe for asynchronous systems.

n

Now has several core functions

• Reliable order-preserving “Optimized” Message transport

(based on performance measurement) in heterogeneous

multi-link fashion with TCP, UDP, SSL, HTTP, and will add GridFTP

• General publish-subscribe including JMS & JXTA and

support for RTP-based audio/video conferencing

• Distributed XML event selection using XPATH metaphor • QoS, Security profiles for sent and received messages

Laudable Features of NaradaBrokering

n

Is

open source

http://www.naradabrokering.org

n

W

ill have

client

“plug-in” as well as standalone brokers

n

Will have a

discovery service

to find nearest brokers

n

Does

tunnel

through most firewalls without requiring

ports to be opened

n

Supports

JXTA, JMS

(Java Message Service) and more

powerful native mode

n

Transit time

< 1 millisecond

per broker

n

Will have

setup

and

broker network administration

NaradaBrokering Naturally Supports

n

Filtering

of events to support different client

requirements (e.g,. PDA versus desktop, slow lines,

different A/V codecs)

n

Virtualization

of addressing, routing, interfaces

n

Federation

and

Mediation

of multiple instances of Grid

services as illustrated by

• Composition of Gridlets into full Grids (Gridlets are single

computers in P2P case)

• JXTA with peer-group forming a Gridlet

n

Monitoring

of messages for Service management and

general autonomic functions

n

Fault tolerant

data

transport

Grid Messaging Substrate

Consum

er Service

SOAP+HTT GridFT

RTP ….

Messaging Substrate Consum

er Service

Standard client-server style communication.

Substrate mediated

communication removes transport protocol

dependence.

SOAP+HTT GridFT

RTP ….

Any Protocol satisfying QoS

Heterogeneous Routing in NB

n Mediation in Cellular

Grid using NB as interface agent

• Build Virtual Private Grid Gridlets Satellit UDP Firewal HTTP Dial-u Filter A B1 Hand-Hel Protocol Fas Link

Software Multicast B2

B3

NB Brokers _{Client Filtering}

Architecture of Message Layer

n Need to optimize not only routing of particular messages but

classic publish/subscribe problem of integrating different

requests with related topics (subscribe to sports/basketball/lakers

and sports)

n Related to Akamai, AOL … caching and Server optimization

problem

1-> N Grid Clients

Hypercube o

NB Brokers (logical not physical)

Autonomic Services

n In a Web (Grid) Service architecture, the state of any service is defined by its initial condition and all the messages (including ordering) that it receives

• This how shared event model of collaboration works

n This is a “Finite State Change” model analogous to saving file and “undo” command in many editors

n NB plus a robust store can “guarantee” to save all these messages for (all) services

n This allows one to build both "autonomic data transport" and

"autonomic services" since these services can sustain packet losses in transport and can also sustain failures of apps/brokers

• archived messages (previous invocations, published events etc) can

be retransmitted to reconstruct state at the service or to correct a transport error.

n Further anomalies in message traffic (such as a publisher or

subscriber are silent) can be detected by NB and signal problems n We are building examples of both scenarios using GridFTP as our

data transport example

Collaborative SVG Web Service

n SVG is W3C 2D Vector Graphics standard and is interesting for

visualization and as a simple PowerPoint like application

• Further SVG is built on W3C DOM and one can generalize results

to all W3C DOM-based applications (“all” in future?)

n Apache Batik SVG is Java and open source and so it is practical

to modify it to explore

• Real Applications as a Web Service • Collaboration as a Web Service

• MVC model and web services with implications for portlets

n We use NaradaBrokering and XGSP to control collaboration;

support PDA Cell-phone and desktop clients; are restructuring Batik as MVC Web Service

• Good progress in all areas see

• http://www.svgarena.org for SVG Games

Web Service Model for Application Development

W3C DOM User Interface W3C DOM Raw (UI) Events

Application as a Web service

W3C DOM Semantic Events Data

User Facin Ports

Resource Facing Ports

Events as Message s Rendering as Messages View Control Model Narad Brokering

Interrupts in traditional monolithic applications becom “real messages” not directly method calls

Natural for collaboration and universal access

Collaborative SVG As A Web Service

Collaborative SVG Chess

Game in Batik Browser

Players

XGSP Web Service MCU Architecture

SIP H323 Access_Grid Native_XGSP Admire

Gateways convert to uniform XGSP Messaging

High Performance (RTP and XML/SOAP and ..

Media Servers Filters Session Server XGSP-based Control NaradaBrokerin g All Messaging

Use Multiple Media servers to scale to many codecs and many versions of audio/video mixing

NB Scales a distributed

We Services

Polycom, Access Grid

and RealVideo views of

NaradaBrokering Communication

n Applications interface to NaradaBrokering through

UserChannels which NB constructs as a set of links between NB Brokers acting as “waystations” which may need to be

dynamically instantiated

n UserChannels have publish/subscribe semantics with XML topics n Links implement a single conventional “data” protocol.

• Interface to add new transport protocols within the

Framework

• Administrative channel negotiates the best available

communication protocol for each link

n Different links can have different underlying transport

implementations

• Implementations in the current release include support for

TCP,UDP, Multicast, SSL, RTP and HTTP.

• Supports communication through proxies and firewalls such as

Pentium-3, 1GHz, 256 MB RAM

100 Mbps LAN

JRE 1.3 Linux

hop-3 0 1 2 3 4 5 6 7 8 9 100 1000 Transit Delay (Milliseconds)

Message Payload Size (Bytes)

Mean transit delay for message samples in NaradaBrokering: Different communication hops

hop-2

0 1 0 2 0 3 0 4 0 5 0 6 0 0 20

0 400 600 800 1000 1200 1400 1600 1800 2000 Delay (Milliseconds)

Packet Number

Average delays per packet for 50

video-clients_{NaradaBrokering Avg=2.23 ms, JMF Avg=3.08}

ms

NaradaBrokering-RTP

0 1 2 3 4 5 6 7 8 0 20

0 400 600 800 1000 1200 1400 1600 1800 2000

Jit

ter

(Milliseconds)

Packet Number

Average jitter (std. dev) for 50 video clients.

NaradaBrokering Avg=0.95 ms, JMF Avg=1.10

ms

NaradaBrokering-RTP

Collaboration and Web Services

n

Collaboration

has

• Mechanism to set up members (people, devices) of a

“collaborative sessions”

• Shared generic tools such as text chat, white boards,

audio-video conferencing

• Shared applications such as Web Pages, PowerPoint,

Visualization, maps, (medical) instruments ….

n

b)

and

c)

are “just shared objects” where objects

could be Web Services but rarely are at moment

• We can port objects to Web Services and build a general

approach for making Web services collaborative

n

a)

is a “Service” which is set up in many different

Shared Event Collaboration

n All collaboration is about sharing events defining state changes

• Audio/Video conferencing shares events specifying in compressed form

audio or video

• Shared display shares events corresponding to change in pixels of a frame

buffer

• Instant Messengers share updates to text message streams

• Microsoft events for shared PowerPoint (file replicated between clients) as

in Access Grid

n Finite State Change NOT Finite State Machine architecture n Using Web services allows one to expose updates of all kinds as

messages

• “Event service” for collaboration is similar to Grid notification service

and we effectively define SDE’s (service data elements) in OGSI

n Group (Session) communication service is needed for the

delivery of the update events

Global-MMCS 2.0 (1) XGSP MCU

n We are building an open source protocol independent Web

Service “MCU” which will scale to an arbitrary number of users and provide integrated thousands of simultaneous users

collaboration services.

n We will deploy it globally and hope to test with later this year. n The function of A/V media server will be distributed using

NaradaBrokering architecture.

• Media Servers mix and convert A/V streams

n Open XGSP MCU based on the following open source projects

• openh323 is basis of H323 Gateway

• NIST SIP stack is basis of SIP Gateway

W Displa y W Viewe r WS Displa y WS Viewe r Even (Message Service Master W Displa y WS Viewe r

Web Service Message Interceptor

Collaboration as a W Set up Session with XGSP

Application o Content source WSD L Web Service F I U O F I R O

W Displa y W Viewe r WS Displa y WS Viewe r Even (Message Service Master W Displa y WS Viewe r

Collaboration as a W Set up Session with XGSP

We Servic e F I U O F I R O

Shared Input Port (Replicated WS) Collaboration

XGSP Conference Control

Framework Components

n

User session

management

• User session management supports user sign-in, user

create/terminate/join/leave/invite-into XGSP sessions.

n

Application Session

Management

• XGSP application session management provides the services

to A/V and data application endpoints and communities, controlling multipoint A/V RTP and data channels.

n

Floor

Control

• Floor control manages the access to shared collaboration

Performance Test : GlobalMMCS1.0

n

We conducted extensive performance tests on audio

and video servers.

n Video:

• The test shows that our video server is capable of supporting

300 clients if there is only one video sender.

• Video Server Machine : 1.2GHz Intel Pentium III dual CPU,

1GB MEM, RedHat Linux 7.3

n Audio:

• Our tests show that audio server can support 5 concurrent

sessions (250 participants in total) without any packet droppings.

• Audio Server Machine: 2.5GHz Pentium 4 CPU, 512MB memory,

Windows XP machine

Global-MMCS 2.0 (2) Portlets

n

Collaboration clients will be built into portlets by

creating

Java Applet or ActiveX controls

for the

non-HTML clients and adding them into non-HTML pages.

n

A collaboration portlet opens local services for XGSP

application session management and floor control.

• Node Manager portlet invoke the service to control local

portlets

n

Apache Jetspeed

seems good open source technology

supporting this model

n

Portlets such as

Access Grid portlet

(really a VIC portlet)

Multicast

Multi-stream AG Portlet

n

Java applet supports

multicast AG with

multiple streams

n

In Jetspeed, easiest to

Workflow GCEs and Problem Solving

Environments (PSEs)

•

There is some confusion between fields of workflow

(Grid Computing Environments GCE) and PSEs

•

To extent PSEs “just” allow manipulation of “nuggets”,

they are indistinguishable from a domain specific GCE

•

They are distinct if they support intra nugget operations

such as

– Integration of mesh and simulation – Closely coupled code linkage

– Generation of code from high level interface like Mathematica

Web Services as a Portlet

•

Each

Web Service

naturally has a

user interface

specified as “just

another port”

– Customizable for universal access

•

This gives each Web Service a

Portlet

view specified (in XML as

always) by

WSRP

(Web services

for Remote Portals)

•

So component model for resources

“automatically” gives a

component

model for user interfaces

– When you build your

application, you define portle

at same time

Application o Content source WSD L Web Service S R W P

Application as a WS

General Application Port Interface with other We Services

User Face o Web Servic

WSRP Ports define WS as a Portlet

Web Services have other ports (Grid Service) to be

Online Knowledge Center built from Portlets

• Web Services

provide a

component model

for the middleware (see large “

common

component architecture

” effort in Dept. of

Energy)

• Should match each WSDL component with

a corresponding user interface component

• Thus one “must use” a

component model

for the portal

with again an XML

specification (

portalML

) of portal

component

Portlet Portlet Portlet Portlet

XML

RSS, OCS, or other Local or remote

HTML

Local files

JSP or VM

Local templates WebPageRemote HTML

Portlet

Portlets

User

implemented using Portal API

Portlets

Data

PortletControlle

r PortletControlle_r

Screen Manager HTML PSML PortletContro l EC S JSP template EC

S EC_S EC

S ECS

EC

S ECS ECS

ECS Root to HTML

EC S

Turbine Servlet Jetspeed

Portlets and Portal Stacks

• User interfaces to Portal services (Code

Submission, Job Monitoring, File

Management for Host X) are all managed as

portlets.

• Users, administrators can customize their portal

interfaces to just precisely the services they want.

Core Grid Services User facing Web

Service Ports

IU and OGCE Portal Architecture

Client s (P ure HT ML, Java Applet ..) Aggregat ion and Rendering Jetspeed Internal Services Portlet Class IFramePortlet Portlet Class VelocityPortlet Portlet Class JspPortlet Portlet Class

WebForm Gateway(IU) Web/Griservice

Web/Gri service Web/Gri service Computing Data Stores Instruments GridPort Texas (Java) COG Kit

Clients Portal Portlets Libraries Services Resources Loca

Portlets

Remot or Prox Portlets

Emphasis _{from other projects}Largely take

(Jetspeed)

Jetspeed Computing Portal: Choose Portlets

4 available portlet

Choose Portlet Layout

Choose 1-column Layout

Lists user files on selected host, noahsark.

File operations include Upload, download, Copy, rename, crossload

Tabs indicate available

portlet interfaces.

File

Sample page with several portlets:

Provide information about application

and

host parameters

Select application to edit