1. Griffin Update: Toward an Agile, Predictive Infrastructure Anthony D. Joseph UC Berkeley http://www.cs.berkeley.edu/~adj/ Sahara Retreat, June 2003

2. 2 Outline Griffin – Motivation – Goals – Components Tapas Update Tapestry/Brocade Update REAP/MINO Update

3. 3 Near-Continuous, Highly-Variable Internet Connectivity Connectivity everywhere: campus, in-building, satellite… – Projects: Sahara (01-), Iceberg (98-01), Rover (95-97) Most applications support limited variability (1% to 2x) – Design environment for legacy apps is static desktop LAN – Strong abstraction boundaries (APIs) hide the # of RPCs But, today’s apps see a wider range of variability – 3  5 orders of magnitude of bandwidth from 10's Kb/s  1 Gb/s – 4  6 orders of magnitude of latency from 1  sec  1,000's ms – 5  9 orders of magnitude of loss rates from 10 -3  10 -12 BER – Neither best-effort or unbounded retransmission may be ideal – Also, overloaded servers / limited resources on mobile devices Result: Poor/variable performance from legacy apps

4. 4 Griffin Goals Users always see excellent (  local, lightly loaded) application behavior and performance – Independent of the current infrastructure conditions – Move away from “reactive to change” model – Agility: key metric is time to react and adapt Help legacy applications handle changing conditions – Analyze, classify, and predict behavior – Pre-stage dynamic/static code/data (activate on demand) Architecture for developing new applications – Input/control mechanisms for new applications – Application developer tools

5. 5 Griffin: An Adaptive, Predictive Approach Continuous, cross-layer, multi-timescale introspection – Collect & cluster link, network, and application protocol events – Broader-scale: Correlate AND communicate short-/long-term events and effects at multiple levels (breaks abstractions) – Challenge: Building accurate models of correlated events Convey app reqs/network info to/from lower-levels – Break abstraction boundaries in a controlled way – Challenge: Extensible interfaces to avoid existing least common denominator problems Overlay more powerful network model on top of IP – Avoid standardization delays/inertia – Enables dynamic service placement – Challenge: Efficient interoperation with IP routing policies

6. 6 Some Enabling Infrastructure Components Tapas network characteristics toolkit – Measuring/modeling/emulating/predicting delay, loss, … – Provides micro-scale network weather information – Mechanism for monitoring/predicting available QoS REAP protocol modifying / application building toolkit – Introspective mobile code/data support for legacy / new apps – Provides dynamic placement of data and service components – MINO E-mail application on OceanStore / Planet Lab Tapestry, Brocade, and Mobile Tapestry – Overlay routing layer providing efficient application-level object location and routing – Mobility support, fault-tolerance, varying delivery semantics

7. 7 Outline Griffin – Motivation – Goals – Components Tapas Update Tapestry/Brocade Update REAP/MINO Update

8. 8 Tapas Accurate modeling and emulation for protocol design – Very difficult to gain access to new or experimental networks – Delay, error, congestion in IP, GSM, GPRS, 1xRTT, 802.11a/b – Study interactions between protocols at different levels Goal: Create models/artificial traces that are statistically indistinguishable from real network traces – Such models have both predictive and descriptive power – Better understanding of network characteristics – Can be used to optimize new and existing protocols Tapas: Novel data preconditioning-based analysis – More accurately models/emulates long-/short-term dependence effects than classic approaches (Gilbert, Markov, HMM, Bernoulli)

9. 9 Tapas Update Domain analysis tool – Chooses most accurate model for a metric Markov-based Trace Analysis, Modified hidden Markov Model New Tapas-based link simulator – Complete reimplementation of Wsim – Enables quick and repeatable testing of new apps Tapas talk this afternoon

10. 10 Domain Analysis: Choosing the Right Network Model Collect empirical packet trace: T = {1,0}* – 1: corrupted/delayed packet, 0: correct/non-delayed packet Create mathematical models based on T Challenge: domain analysis – which model to use? – Gilbert, HMM, MTA, M3 have different properties Algorithm (applied to Gilbert, HMM, MTA, M3): – Collect traces, compute exponential functions for lengths of good and bad state and compute 1’s density of bad state – For a given density, determine model parameters and optimal model (best Correlation Coefficient) Sigmetrics 2003 paper network model artificial network metric trace trace analysis algorithm real network metric trace

11. 11 Domain Analysis Experiment Create artificial network environment with varying bad state densities (generate synthetic reference traces) For each trace: – Create classical and data preconditioning models – Generate artificial traces from models – Plot error and error free distribution – Calculate Correlation Coefficient (CC) between distributions of reference and artificial traces Optimal model for a given set of properties is the one with the highest CC value – Plot optimal model as a function of the good and bad state exponential values: Domain of Applicability Plot

12. 12 Domain of Applicability Plot, L den = 0.2

13. 13 Domain of Applicability Plot, L den = 0.7

14. 14 Tapestry/Brocade Starting point is Tapestry – Distributed Object Location and Routing (DOLR) overlay network Extend Tapestry with unique, powerful routing functions – SLA-compatible efficient wide-area routing – Rapid, scalable mobility support – Rapid fault route-around using pre-computed backup routes – Monitoring, measurement, and analysis entry point

15. 15 Tapestry/Brocade Update Major push to improve Tapestry reliability – Pre-computed backup paths enable near- instantaneous fail-over (3 paths/router entry) – Improved Patchwork network link monitoring Now ready for integration with link prediction support – Improved repair algorithms to handle long-term faults Building new applications – SpamWatch (Middleware 2003 paper) Summer focus on inter-domain Brocade routing

16. 16 Improved Tapestry Fault-Tolerance

17. 17 REAP/MINO Introspective code / data migration in 3-tier hierarchies – Distributes server load, empowers limited devices – Provides illusion of high connectivity Combines static trace analysis w/ dynamic monitoring of clients to predict appl’n / communication behavior – Identify and optimize code/data placement – Pre-stage statically/dynamically generated components – Explore various granularities of code & data migration – Predict costs using multiple criteria MINO E-mail OceanStore application – Basis for exploring code/data migration choices

18. 18 REAP/MINO Update Code migration work is mostly complete, now focused on data migration – Understanding how users access data and how they move, so that we can better place/cache data Collecting user mobility / data access traces – Web proxy traces from NLANR (access clustering) – EECS IMAP server traces (user location, access info) – College campus NFS-level traces from a login server (used for e-mail reading) Built cooperative caching simulator – Exploring multi-criteria optimization: frequency of access, number of users accessing, $ paid/user, etc.. – Cache refill can leverage link predictor information

19. 19 Recent Griffin Progress Summary Tapas: Network modeling and analysis – Thesis: Almudena Konrad, “TAPAS: A Research Paradigm for the Modeling, Prediction and Analysis of Non-stationary Network Behavior,” (PhD, expected August 2003) – Simulator almost ready for release – Publications Konrad, A.; Joseph, A. D., Choosing an Accurate Network Path Model, In Proc. Of SIGMETRICS 2003, June, 2003. Konrad, A.; Zhao, B. Y.; Joseph, A. D.; Ludwig, R., A Markov-Based Channel Model Algorithm for Wireless Networks, ACM Wireless Networks, vol. 9, num. 3, May, 2003.

20. 20 Recent Griffin Progress Summary Tapastry / Brocade: – Robustness fixes to Tapestry, lots of measurements – Publications Zhao, B. Y.; Huang, L.; Stribling, J.; Rhea, S. C.; Joseph, A. D.; Kubiatowicz J. D., Tapestry: A Resilient Global-scale Overlay for Service Deployment. To appear in IEEE JSAC, Fall 2003. Zhou, F.; Zhuang, L.; Zhao, B.; Huang, L.; Joseph, A. D.; Kubiatowicz, J., Approximate Object Location and Spam Filtering on Peer-to-Peer Systems, In Proc.of ACM Middleware 2003, June, 2003. REAP/MINO – Simulator developed, lots of traces collected – Beginning analysis phase

21. Griffin Update: Toward an Agile, Predictive Infrastructure Anthony D. Joseph UC Berkeley http://www.cs.berkeley.edu/~adj/ Sahara Retreat, June 2003