Docstoc

Pervasive Services Infrastructure

Document Sample
Pervasive Services Infrastructure Powered By Docstoc
					Pervasive Services Infrastructure (Ψ)
             Dejan S. Milojicic

               HP Labs, Palo Alto
                   3/15 2001
     http://www.hpl.hp.com/research/itc/csl/pss/psi


                  3/18/2001 Y, Pervasive Service Infrastructure   1
                       Ψ’s Strategic Direction



Internet           Wireless

                                     Global eCommerce to reach $6.8 trillion by
                                      2004 (Forrester Research)
            Service                  Internet Service Gateway market to reach 25M
           Providers
                                      US homes by 2005, worth $5B (Parks Assoc.)
                                     Commerce over mobile phones in W. Europe
                                      rise to $37.7B in 2004 ($51.2M in 1999, IDC)
                                     Worldwide shipments of handheld computers
                                      will surpass 5.7M, 47% increase (Dataquest)

                              3/18/2001 Y, Pervasive Service Infrastructure      2
          What is Disruptive in this Space

 General
      pervasive (ubiquitous, invisible) deployment of computers, smart spaces
      scale: large number of devices (localized) & services
 User perspective
      user interfaces, user intent
      context awareness
      diversity of services and clients
      connectivity
 Developer perspective
      versioning and maintaining products (scale thereof)
      variation in network speed (wireless → wired; wired → memory)
      the price of each component (reduced cost)
      service interoperability (everything available on the Internet)


                               3/18/2001 Y, Pervasive Service Infrastructure     3
                Tough Problems Today

   Wireless speed (need higher bandwidth, lower latency)
   Unreliable connections (missing disconnected support)
   Lack of apps and app development tools
   Lack of solid infrastructure (database, synchronization)
   Non-scalable solutions
   Non-trusted solutions (lack of end-to-end security)
   Lack of quality displays & U/I
   Few users (innovators and early adopters)
   Protocols evolving (WAP, XML implementations, APIs)
   User experience (clumsy handheld, poor I/F)
   Viability (fractured embedded OS space)

                          3/18/2001 Y, Pervasive Service Infrastructure   4
           Relaxing Assumptions for 2004
                     (not all problems are Y ’s)

   Ubiquitous connectivity (3G+)
   End-to-end security (Wireless VPN)
   Device interoperability (Lucent & Novell, Motorola & Lucent, etc)
   Better battery lifetimes (piezoelectric, solar, bioelectric, etc.)
   Handheld, phones, others,… converged
   Billing, customer care → commodity services
   Voice U/I (many already work on)
   Higher quality color displays
   New Web apps built from ground up
   Middleware solutions adapt access to existing apps
   Wireless Web crosses the chasm
   Consolidation of embedded OS space

                            3/18/2001 Y, Pervasive Service Infrastructure   5
          What are Remaining Problems

   Scalability (localized)
   Service access and adaptivity in dynamic environments
   Administration (automatic, transparent, dependency aware)
   One-size fits all (embedded system software solutions)
   Interoperability (APIs, service brokers, etc.)
   Performance e.g. load, latency, caching
   Maintaining true end-to-end security guarantees




                         3/18/2001 Y, Pervasive Service Infrastructure   6
                                       Scenario
   Scene: Helsinki airport, year 2004. Jane and Dave are on a business trip to present a new
   product to overseas customers. Dave worked the whole weekend at home on an updated
   presentation. They are stepping off of the plane.

Jane: Relax Dave, everything is going to be fine. Your presentation is in good shape. Just
   fix the typo in the CTO’s name and tune the figures,
Dave: Fix the title! Tune the figures! Are you out of your mind?! I only have the
   presentation on my phone. I’d need to download and fix it on my laptop. We don’t
   have time for that before the meeting.
Jane: Yes, you can. I can do it on my new phonDA.
Dave: You never told me you got a new, more powerful phonDA
Jane: No, it’s the same as yours, NK234yP, but I’ve updated the software.
Dave: No way, my presentation is 4G. It can’t fit on any phonDA
Jane: Look, I download your presentation ... Here’s the slide you want, … ok… the CTO’s
   title is fixed, … now let’s see the figures… it’ll take longer… ok, here are figures,
   move the 2003 forecast up a little, .. perfect.

                                 3/18/2001 Y, Pervasive Service Infrastructure                 7
                         Scenario, cont.

Dave: How did you do it?!! Look, I’m downloading it on my phonDA and it
   takes forever… even worse, now it won’t start.
Jane: You see those small bumps on the walls, ceiling, and floor. These are
   embedded servers. Software on my phonDA offloads networking, memory,
   processing, storage - you name it - to these servers.
Dave: Sigh…
Jane: It is called Psi; it’s very simple. Now with respect to these bumps on
   walls, … hold on a second… my daughter is paging me, first things first….
   Yes honey, … I forgot to sign your homework … ooops …
Dave: Can your magic Psi help with this too?
Jane: Sure it can, see, here it will download her application from her school – it
   adapts to my phonDA since the app is designed for desktops– it’s really
   downloaded on a server there. Looks correct, ok signed!


                            3/18/2001 Y, Pervasive Service Infrastructure            8
                         Scenario, cont.

Dave: Wow, you are a wizard, installing all these applications, adapting
   them to your small phonDA. How long did it take you?
Jane: Not a second, Psi does it all, it adapts, it downloads services on
   demand.

                                          Sigh                              Ψ



                            3/18/2001 Y, Pervasive Service Infrastructure       9
Pervasive Servers & Clients
                               ..servers…



                                                               IDCs


                                                       infrastructure

                          ..clients…                              Ψ

      ..embedded devices and sensors…                     embedded
       3/18/2001 Y, Pervasive Service Infrastructure               10
      Conceptual Layers


     Services and Applications

         Middleware Layers
     E-Speak, Jini, CORBA, ACL
                  Ψ platform:
match service requirements to client resources

           Local OSes & JVMs
            Client Handheld Device
          (and Infrastructure Servers)
            3/18/2001 Y, Pervasive Service Infrastructure   11
                                 Ψ Vision

  Adapt any service to any client (anywhere, anytime)

 General assumptions
      clients will always be diverse
      clients will always be less powerful than desktops
      there will be a large scale of devices in the future
      services will be accessible on demand from anywhere

 Underlying Y assumptions:
      service adaptivity required
      three-tier model
      service splitting
                           3/18/2001 Y, Pervasive Service Infrastructure   12
              Ψ, Important Research Questions

 How to enable users to exploit the new pervasive computer infrastructure
      How to seamlessly offer more services to more clients anytime anywhere?
      How to avoid installing and administering increasing number of computers in a
       pervasive infrastructure?
      What are new abstractions and algorithms for computation, communication,
       storage, user interface → how to write and run new apps?



  It is not the strongest of the species that survive, nor the most intelligent, but
  the ones most responsive to change - Charles Darwin

                                3/18/2001 Y, Pervasive Service Infrastructure          13
             Adaptive Offloaded Services

 Goal: increase scalability & performance of (mobile) service delivery
  to resource-poor devices and enable where not currently possible
 Key research questions
      Splitting service between client and mid-point (intermediate) servers
      Dynamic adjustment of services: device size, load, roaming
      Masking performance implications of splitting
 Required infrastructure topics
      Distributed run-time support - mid-point(s) and client
      Sharing, administration, and security at mid-points




                            3/18/2001 Y, Pervasive Service Infrastructure      14
                   Adaptive Offloaded Services
                            Illustration

                                                                                                 1110
                                                                                                 0101
                                                                                                 1110

                     Internet
                                                                                                 0101



                                                                                1110
                                                                                0101
                                                                                1110
                                                                                0101

010101
111010
010101                                                                                                        1110
111010                                                                                                        0101
    0    IDC                    Embedded server
                                                                                                              1110
                                                                                                              0101

                                     dependent                                                    Which?
                                      execution


Back-end service
   execution
                                                                                       Home, office, shop, etc.
                                3/18/2001 Y, Pervasive Service Infrastructure                                        15
       Adaptive Offloaded Services, Cont.

 Splitting services
      Monitoring (resources, execution, objects)
      Offloading (migration mechanisms, trigger & placement policies)
      Service splitting (interaction metrics, graph partitioning, 3-v. 2-tier)
 Adaptive Offloading
      Adapting to devices (very constrained, different resources)
      Adapting to load (multiple services, scalability)
      Adapting to mobility (roaming, migration of offloaded state)
 Performance
      Policies (account for overhead, interaction metric, stability)
      Mechanisms (offloading overhead, disruption of offloading - lazy)

                             3/18/2001 Y, Pervasive Service Infrastructure        16
                   The Java Heap Page Working Set
                     of Embedded Caffeine Mark
                   250
Usage (4k Pages)
 Accum. Memory




                   200
                                                                                                         Unt ouched Live
                                                                                                         Object s
                   150
                                                                                                         Thread St ack
                                                                                                         Touched
                   100
                                                                                                         Heap Object s
                   50
                                                                                                         Touched


                    0
                         0k   50k   100k     150k          200k          250k              300k   350k

                               Time (Instructions Executed)
                                           3/18/2001 Y, Pervasive Service Infrastructure                                   17
                    Services on Demand

 Goal: zero-administration and -installation of mobile clients
 Key research questions
       dynamic service composition and deployment
        on zero-installed devices
       de/re-coupling of user’s service context between devices
       tolerance to disconnection of services
 Required infrastructure topics
       service brokers and infrastructure
       device encapsulation APIs




                            3/18/2001 Y, Pervasive Service Infrastructure   18
              Services on Demand Illustration
            Day 1, John in Sydney




                                                                                     Secured Storage
                           Internet                                                      Provider




Day 2, John in Montreal                                                      Service broker                    ASP


 Back-end service
    execution                                                                                          ASP

                                                                                    Home, office, shop, etc.
                                    3/18/2001 Y, Pervasive Service Infrastructure                                    19
                        Loading Overheads

              Local          SoD Loading (s)                                     Loaded URLs      Java Class
  Service
             Load (s)   Intranet      Wave                Dialup                Cnt   Size (B)   Cnt   Size (B)
calculator     0.1        ~0           +0.1                 +4.7                2      9520       2     9520
calendar       0.2        ~0           +0.4                 +4.7                4     12952       4    12952
editor         0.2        ~0           +0.3                 +5.2                7     15885       7    15885
game           0.2        ~0           +0.3                 +6.1                9     18739       9    18739
agenda         0.2       +0.1          +0.7                 +8.4                1     35360      12    59846
ftp            0.6       +2.1          +3.2                +20.9                3     107725     22    142155
mail           2.5       +2.7         +15.1              +129.4                 1     675046     138   365574


                                3/18/2001 Y, Pervasive Service Infrastructure                                  20
                  Services on Demand, Cont.
 Service Composition/Deployment on Zero-Installed Devices
       service characterization (naming, keyword, extension, requirements)
       service discovery (brokers, resource matching, scalability)
       deployment (transparent downloading & caching, sandboxing for service inter-
        communication
 De/re-Coupling of Service Context Between Devices
       remote storage (clients are volatile, access any time anywhere, user/service
        transparency, stackable client RFS, synchronization/encryption)
       context information (per service state, preferences, expressed as files)
       Adaptation to different client (resource re-evaluation, select service)
 Tolerance to Disconnection of Services
       client side cache (static content: URLs, classes; user files/data; user/service network
        interactions
       reconnection (pluggable reconciliation per service/file, client-initiated


                                  3/18/2001 Y, Pervasive Service Infrastructure               21
                Architectural Principles

 Minimal changes (extensions) to the underlying client sys SW
 Minimal changes to services (mainly splitting)
 Use existing mid-point servers in the infrastructure
 Psi will rely on standard communication protocols (IP, WAP, etc)
  however, it will have the ability to support extension to new ones
 Heterogeneity of underlying OSes and implementation of JVMs




 The primary metric will be enabling new services and
  better use of resource-constrained devices

                         3/18/2001 Y, Pervasive Service Infrastructure   22
                 Implementation Details

 Prototype environment
      Jornada Pocket PC
      Java VM, Chai VM
      wireless 802.11
      Linux PCs for infrastructure servers
 Services
      tools: editors (text/image)
      visualization
      PIM: calendars, mail, calculators, stock analysis
      interactive games
      etc.

                           3/18/2001 Y, Pervasive Service Infrastructure   23
                    Technical Approach

 Develop a Y architecture & prototype implementation
 Experimental, quick prototyping, 3-month increments
 Spiral approach
      Incrementally demonstrate new functionality
      Uncover new key questions
 Engage partners at incremental 3-month phases
      HPL, HP, university, industry (external)
 Patent/publish as we progress




                           3/18/2001 Y, Pervasive Service Infrastructure   24
                               Showcases

 Adaptive offloaded services
  initial mobile app splitting, evaluate split advantage (scale/perf/power)
       app which is too resource intensive for existing devices
       app which is interactive → mobile editing and interactive gaming
       app which is data intensive → mobile multimedia and visualization
 Service-on-Demand
   service broker lookup, automatic download, disconnected mode (cache/remote)
      app that can work disconnected (runs part of the service locally)
      app that accesses user data from a server
      app in two different flavors, to adapt to Palm constrained resources



                               3/18/2001 Y, Pervasive Service Infrastructure   25
                      Current Insights

   Opportunity for memory offloading
   Complex interactions in Java objects
   Transparent remote storage: class interposition, bytecode editing
   Disconnection: caching of some class client/service interactions
   Reconciliation through plugins




                         3/18/2001 Y, Pervasive Service Infrastructure   26
                         Team Members

 Resource-Constrained Devices, team lead
  OS, JVM, consumer products (Sony)
 Team member
  HA, distributed systems (SRI, Informix, Oracle)
 Services-on-Demand, team lead
  OS, JVM (Bull, OSF RI)
 Dejan Milojicic, (myself) Psi PM
  OS, distributed systems & agents (Belgrade RI, OSF RI)




                             3/18/2001 Y, Pervasive Service Infrastructure   27
                             Competition
 Telcos
      AT&T, Motorola, Ericsson
 Consumer electronics companies
      Sony, Philips, etc. (HAVI)
 Traditional computer systems companies
      Sun (Jini, embedded/real-time Java); Microsoft (.NET, Windows CE)
       IBM (pervasive computing, embedded tools)
 Data base companies
      Oracle, Sybase
 Many startups
      StreamTheory, OmniShift, Transvirtual, M2Verticom, WirelessKnowledge
       Many universities, government
 Numerous competition, but huge business & innovation opportunity

                             3/18/2001 Y, Pervasive Service Infrastructure    28
                              Related Work
 Adaptability & Offloading
      UW (Portolano) Active Fabrics + service infrastructure
      Berkeley’s (Endeavor) Ninja + general service offloading
      Palm, WinCE + offloading to shared mid-points
      OSGi + general purpose + distributed; UW Kimera + dynamic service split
      CMU Odyssey, UIUC Active Spaces + service splitting
      MIT Oxygen SW environment + resource awareness
 Data storage & resource management
      CMU Coda & Odyssey + reconciliation framework
      UCLA File mobility + trust for services; WebFS + more than web browsing
      OSGI + mobility; Java OS JDI & JPI + API signatures
 Java platform
      JavaOS + service infrastructure; PocketLinux + disconnected + other OSes
      JNT, but not just a terminal; WebOS + service brokers & trust
                               3/18/2001 Y, Pervasive Service Infrastructure      29
                 What Ψ is Not About

   Client HW
   HW (re)configuration
   Service design & implementation
   New protocols
   New programming languages
   General purpose wireless
   Device location technologies
 General purpose OS development
 Fault-tolerance
 Hard real-time


                        3/18/2001 Y, Pervasive Service Infrastructure   30
                                Summary

 Intersection of services, the Internet, and wireless:
       adapt services to existing client resources
       automate service deployment
       reason about new, disruptive technologies in pervasive computing
 Potential long-term area of research, ties in
       servers (IDCs) & client pervasive infrastructures
       academia, government, big & small companies
       different markets
 A lot of opportunities for contribution; risk-aware, leverage potential
 It is fun to do these things! Immediate benefit obvious


                            3/18/2001 Y, Pervasive Service Infrastructure   31
      Pervasive Computing Landscape

                                                       application model
             user/agent model
                                                     (information access
               (intelligence)
                                                        & presentation)



                                      global
                         task                           location             context        information
          nomadic       mobility
                                       data
                                                     transparency          management                        security
                                    placement                                               architecture
                                                                                                               mgmt
 user
studies
                                                Y                                              XML
                                                                                                              pricing/
                                                                                                             charging
                                                                                               UDDI
                                                                                                           development
                                                                                               VML
                internet com                                mobile com.                                     adaptivity
                                                                                                …
                     (IP)                                    (non IP)


                                                                                        Y
                      device                                    service
                    connectivity                              connectivity




                                   3/18/2001 Y, Pervasive Service Infrastructure                                         32