Clients Proxy Cache Servers Inter-proxy communication distance Proxy utilization and load balance Size of client population served Proxy Clients
|
|
- Justin Stanley
- 5 years ago
- Views:
Transcription
1 achig for a Better Web Large-Scale Web achig ad otet Delivery Jeff hase PS 22: Distributed Iformatio Systems Fall 2000 PerformaceisamajorcoceritheWeb Proxy cachig is the most widely used method to improve Web performace Duplicate requests to the same documet served from cache Hits reduce latecy, etwork utilizatio, server load Misses icrease latecy (extra hops) Hits Misses Misses Iteret liets Proxy ache Servers ache Effectiveess Previous work has show that hit rate icreases with populatio size [Duska et al. 97, Breslau et al. 98] However, sigle proxy caches have practical limits Load, etwork topology, orgaizatioal costraits Oe techique to scale the cliet populatio is to have proxy caches cooperate ooperative Web Proxy achig Sharig ad/or coordiatio of cache state amog multiple Web proxy cache odes Effectiveess of proxy cooperatio depeds o: Iter-proxy commuicatio distace Size of cliet populatio served Proxy Iteret Proxy utilizatio ad load balace liets liets liets Hierarchical aches otet-sharig Amog Peers Idea: place caches at exchage or switchig poits i the etwork, ad cache at each level of the hierarchy. Resolve misses through the paret. origi Web site (e.g., U.S. ogress) upstream Idea: Sice sibligs are close i the etwork, allow them to share their cache cotets directly. dowstream
2 Harvest-Style IP Hierarchies Issues for ache Hierarchies Examples Harvest [Schwartz96] Squid (NLANR) NetApp Netache cliet Idea: multicast probes withi each family : pick first hit respose or wait for all miss resposes. object request object respose query (probe) query respose With IP: query traffic withi families (size ) Iter-siblig IP traffic (ad aggregate overhead) is quadratic with. Query-hadlig overhead grows liearly with. miss latecy Object passes through every cache from origi to cliet: deeper hierarchies scale better, but impose higher latecies. storage A recetly-fetched object is replicated at every level of the tree. effectiveess Iterior cache beefits are limited by capacity if objects are ot likely to live there log (e.g., LRU). Hashig: ache Array Routig Protocol (ARP) Issues for ARP Microsoft Proxy Server GET a-f hash fuctio g-p q-u v-z Advatages. sigle-hop request resolutio 2. o redudat cachig of objects 3. allows cliet-side implemetatio 4. o ew cache-cache protocols 5. recofigurable o way to exploit etwork locality at each level e.g., relies o local browser caches to absorb repeats load balacig hash ca be balaced ad/or weighted with a load factor reflectig the capacity/power of each server must rebalace o server failures Reassigs (/)th of cached URLs for array size. URLs from failed server are evely distributed amog the remaiig - servers. misspealtyad cost to compute the hash I ARP, hash cost is liear i : hash with each ode ad pick the wier. Directory-based: Summary ache for IP ASummary-IP Hierarchy Idea: each cachig server replicates the cache directory ( summary ) of each of its peers (e.g., sibligs). [ao et. al. Sigcomm98] Query a peer oly if its local summary idicates a hit. e.g., Squid cofigured to use cache digests miss Summary caches at each level of the hierarchy reduce iter-siblig miss queries by 95+%. To reduce storage overhead for summaries, implemet the summaries compactly usig Bloom Filters. May yield false hits (e.g., %), but ot false misses. hit Each summary is three orders of magitude smaller tha the cache itself, ad ca be updated by multicastig just the flipped bits. object request object respose cliet query query respose 2
3 Issues for Directory-Based aches Servers update their summaries lazily. Update whe ew etries exceed some threshold percetage. Update delays may yield false hits ad/or false misses. Other ways to reduce directory size? Viciity cache [Gadde/hase/Rabiovich98] Subsettig by popularity [Gadde/hase/Rabiovich97] What are the limits to scalability? If we grow the umber of peers? If we grow the cache sizes? O the Scale ad Performace... [Wolma/Voelker/.../Levy99] is a key paper i this area over the last few years. first egative result i SOSP (?) illustrates tools for evaluatig wide-area systems simulatio ad aalytical modelig illustrates fudametal limits of cachig beefits dictated by referece patters ad object rate of chage forget about capacity, ad assume ideal cooperatio ties together previous work i the field wide-area cooperative cachig strategies aalytical models for Web workloads best traces UW Trace haracteristics A Multi-Orgaizatio Trace Trace UW Duratio 7 days HTTP objects 8.4 millio HTTP requests 82.8 millio Avg. requests/sec 37 Total Bytes 677 GB Servers 244,2 liets 22,984 Uiversity of Washigto (UW) is a large ad diverse cliet populatio Approximately 50K people UW cliet populatio cotais 200 idepedet campus orgaizatios Museums of Art ad Natural History Schools of Medicie, Detistry, Nursig Departmets of omputer Sciece, History, ad Music A trace of UW is effectively a simultaeous trace of 200 diverse cliet orgaizatios Key: Tagged accordig to their orgaizatio i trace ooperatio Across Orgaizatios Treat each UW orgaizatio as a idepedet compay Evaluate cooperative cachig amog these orgaizatios How much Web documet reuse is there amog these orgaizatios? Place a proxy cache i frot of each orgaizatio. What is the beefit of cooperative cachig amog these 200 proxies? Ideal Hit Rates for UW proxies Ideal hit rate - ifiite storage, igore cacheability, expiratios Average ideal local hit rate: 43% 3
4 Ideal Hit Rates for UW proxies Sharig Due to Affiliatio Ideal hit rate - ifiite storage, igore cacheability, expiratios Average ideal local hit rate: 43% Explore beefits of perfect cooperatio rather tha a particular algorithm Average ideal hit rate icreases from 43% to 69% with cooperative cachig UW orgaizatioal sharig vs. radom orgaizatios Differece i weighted averages across all orgs is ~5% acheable hit rate - same as ideal, but does t igore cacheability acheable hit rates are much lower tha ideal (average is 20%) Average cacheable hit rate icreases from 20% to 4% with (perfect) cooperative cachig acheable Hit Rates for UW proxies Scalig ooperative achig Orgaizatios of this size ca beefit sigificatly from cooperative cachig But we do t eed cooperative cachig to hadle the etire UW populatio size A sigle proxy (or small cluster) ca hadle this etire populatio! No techical reaso to use cooperative cachig for this eviromet I the real world, decisios of proxy placemet are ofte political or geographical How effective is cooperative cachig at scales where a sigle cache caot be used? Hit Rate vs. liet Populatio urves similar to other studies [e.g., Duska97, Breslau98] Small orgaizatios Sigificat icrease i hit rate as cliet populatio icreases The reaso why cooperative cachig is effective for UW Large orgaizatios Margial icrease i hit rate as cliet populatio icreases I the Paper.... Do we believe this? What are some possible sources of error i this tracig/simulatio study? What impact might they have? 2. Why are ideal hit rates so much higher for the MS trace, but the cacheable hit rates are the same? What is the correlatio betwee sharig ad cacheability? 3. Why report byte hit rates as well as object hit rates? Is the differece sigificat? What does this tell us about referece patters? 4. How ca it be that byte hit rate icreases with populatio, while badwidth cosumed is liear? 4
5 Trace-Drive Simulatio: Sources of Error What about Latecy?. Ed effects: is the trace iterval log eough? Need adequate time for steady-state behavior to become apparet. 2. Sample size: is the populatio large eough? Is it represetative? 3. ompleteess: does the sample accurately capture the cliet referece streams? What about browser caches ad lower-level proxies? How would they affect the results? 4. liet subsets: how to select to represet a subpopulatio? 5. Is the simulatio accurate/realistic? cacheability, capacity/replacemet, expiratio, latecy From the cliet s perspective, latecy matters far more tha hit rate How does latecy chage with populatio? Media latecies improve oly a few 00mswithidealcachig compared to o cachig. Questios/Issues. How did they obtai these reported latecies? 2. Why report media latecy istead of mea? Is the differece sigificat? What does this tell us? Is it cosistet with the reported byte hit ratios? 3. Why does the magitude of the possible error decrease with populatio? 4. What about the future? What chages i Web behavior might lead to differet coclusios i the future? Will latecy be as importat? Badwidth? Large Orgaizatio ooperatio What is the beefit of cooperative cachig amog large orgaizatios? Explore three ways Liear extrapolatio of UW trace Simultaeous trace of two large orgaizatios (UW ad MS) Aalytic model for populatios beyod trace limits Extrapolatio to Larger liet Populatios UW & Microsoft ooperatio Use least squares fit to create a liear extrapolatio of hit rates Hit rate icreases logarithmically with cliet populatio, e.g., to icrease hit rate by 0%: Need 8 UWs (ideal) Need UWs (cacheable) Low ceilig, though: 6% at 2.M (UW cacheable) A city-wide cooperative cache would get all the beefit Use traces of two large orgaizatios to evaluate cachig systems at medium-scale cliet populatios We collected a Microsoft proxy trace durig same time period as the UW trace ombied populatio is ~80K Icreases the UW populatio by a factor of 3.6 Icreases the MS populatio by a factor of.4 ooperatio amog UW & MS proxies Gives margial beefit: 2-4% Beefit matches hit rate vs. populatio curve 5
6 UW & Microsoft Traces UW & MS ooperative achig Trace UW MS Duratio 7 days 6.25 days HTTP objects 8.4 millio 5.3 millio HTTP requests 82.8 millio 07.7 millio Avg. requests/sec Total Bytes 677 GB N/A Servers 244,2 360,586 liets 22,984 60,233 Populatio ~50,000 ~40,000 Is this worth it? Aalytic Model Iside the Model Use a aalytic model to evaluate cachig systems at very large cliet populatios Parameterize with trace data, extrapolate beyod trace limits Steady-state model Assumes caches are i steady state, do ot start cold Accouts for documet rate of chage Explore growth of Web, variatio i documet popularity, rate of chage Results agree with trace extrapolatios 95% of maximum beefit achieved at the scale of a mediumlarge city (500,000) [Wolma/Voelker/Levy et. al., SOSP 999] refies [Breslau/ao et. al., 999], ad others Approximates asymptotic cache behavior assumig Zipf-like object popularity caches have sufficiet capacity Parameters: λ = per-cliet request rate µ = rate of object chage p c = percetage of objects that are cacheable = Zipf parameter (object popularity) Zipf [Breslau/ao99] ad others observed that Web accesses ca be modeled usig Zipf-like probability distributios. Rak objects by popularity: lower rak i ==> more popular. The probability that ay give referece is to the ith most popular object is p i Not to be cofused with p c, the percetage of cacheable objects. Zipf says: p i is proportioal to /i,forsomewith 0<<. Higher gives more skew: popular objects are way popular. Lower gives a more heavy-tailed distributio. I the Web, rages from 0.6 to 0.8 [Breslau/ao99]. With =0.8, 0.3% of the objects get 40% of requests. acheable Hit Ratio: the Formula N is the hit ratio for cacheable objects achievable by populatio of size N with a uiverse of objects. N = ÿ x = ÿ µ x + λ N x 6
7 Iside the Hit Ratio Formula Iside the Hit Ratio Formula, Part 2 Approximates a sum over a uiverse of objects......of the probability of access to each object x... times the probability x was accessed sice its last chage. What is the probability that i was accessed sice its last ivalidate? = (rate of accesses to i)/(rate of accesses or chages to i) = λnp i /(λnp i + µ) N is just a ormalizig costat for the Zipf-like popularity distributio, which must sum to. is ot to be cofused with N. = ÿ x = ÿ µ x + λn λ x =/Ω i [Breslau/ao 99] 0<< N = ÿ x = ÿ µ x + λ N x Divide through by λnp i. Note: byzipfp i =/i so: /(λnp i )=i /λn Hit Rates From Model Extrapolatig UW & MS Hit Rates acheable Hit Rate Focus o cacheable objects Four curves correspod to differet rate of chage distributios Believe eve Slow ad Mid-Slow are geerous Kee at 500K M These are from the simulatio results, igorig rate of chage (compare to graphs from aalytic model). What is the sigificace of slope? [Graph from Geoff Voelker] Latecy From Model Rate of hage What is more importat, the rate of chage of popular objects or the rate of chage of upopular objects? Straightforward calculatio from the hit rate results Separate popular from upopular objects Look at sesitivity of hit rate to variatios i rate of chage 7
8 Popular docs sesitivity Top curve Upopular low R-of- Issue is miutes to hours Upopular docs sesitivity Bottom curve Popular low R-of- Days to weeks to moth Upopular more sesitive tha popular! ompare differeces i hit rates betwee A, ad B, Rate of hage Sesitivity Hierarchical aches ad DNS What are the implicatios of this study for hierarchical caches ad otet Delivery Networks (e.g., Akamai)? Demad-side proxy caches are widely deployed ad are likely to become ubiquitous. What is the margial beefit from a supply-side DN cache give ubiquitous demad-side proxy cachig? What effect would we expect to see i a trace gathered at a iterior cache? DN iterior caches ca be modeled as upstream caches i a hierarchy, give some simplifyig assumptios. A Idealized Hierarchy N 2 N 2 N Level (Root) Level 2 Assume the trees are symmetric to simplify the math. Igore idividual caches ad solve for each level. Hit Ratio at Iterior Level i N gives us the hit ratio for a complete subtree coverig populatio N The hit ratio predicted at level i or at ay cache i level i over R requests is give by: hits at level requests to level i i h Rpc ( N i i Ni + = = r i ri + hi + the hits for N i (at level i) mius the hits captured by level i+, over the miss stream from level i+ ) Root Hit Ratio Predicted hit ratio for cacheable objects, observed at root of a two-level cache hierarchy (i.e. where r 2 =Rp c ): h N N2 = r N 2 Request Routig Fuctio ƒ(leaf, object, state) Geeralizig to DNs ƒ Iterior aches (supply side reverse proxy ) N I Leaf aches (demad side) N L N L N L N Symmetry assumptio: ƒ is stable ad balaced. 8
9 Hit ratio i DN caches acheable iterior hit ratio Give the symmetry ad balace assumptios, the cacheable hit ratio at the iterior (DN) odes is: NI NL NL N I is the covered populatio at each DN cache. N L is the populatio at each leaf cache. cacheable hit ratio fixed faout N I /N L Iterior hit rates improve as leaf populatios icrease... icreasig N I ad N L --> Iterior hit ratio as percetage of all cacheable requests margial cacheable hit ratio...but, the iterior cache sees a decliig share of traffic. icreasig N I ad N L --> 9
Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5
Morga Kaufma Publishers 26 February, 28 COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Iterface 5 th Editio Chapter 5 Set-Associative Cache Architecture Performace Summary Whe CPU performace icreases:
More informationCMSC Computer Architecture Lecture 11: More Caches. Prof. Yanjing Li University of Chicago
CMSC 22200 Computer Architecture Lecture 11: More Caches Prof. Yajig Li Uiversity of Chicago Lecture Outlie Caches 2 Review Memory hierarchy Cache basics Locality priciples Spatial ad temporal How to access
More information( n+1 2 ) , position=(7+1)/2 =4,(median is observation #4) Median=10lb
Chapter 3 Descriptive Measures Measures of Ceter (Cetral Tedecy) These measures will tell us where is the ceter of our data or where most typical value of a data set lies Mode the value that occurs most
More informationSD vs. SD + One of the most important uses of sample statistics is to estimate the corresponding population parameters.
SD vs. SD + Oe of the most importat uses of sample statistics is to estimate the correspodig populatio parameters. The mea of a represetative sample is a good estimate of the mea of the populatio that
More informationCMSC Computer Architecture Lecture 10: Caches. Prof. Yanjing Li University of Chicago
CMSC 22200 Computer Architecture Lecture 10: Caches Prof. Yajig Li Uiversity of Chicago Midterm Recap Overview ad fudametal cocepts ISA Uarch Datapath, cotrol Sigle cycle, multi cycle Pipeliig Basic idea,
More informationMorgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5.
Morga Kaufma Publishers 26 February, 208 COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Iterface 5 th Editio Chapter 5 Virtual Memory Review: The Memory Hierarchy Take advatage of the priciple
More informationOnes Assignment Method for Solving Traveling Salesman Problem
Joural of mathematics ad computer sciece 0 (0), 58-65 Oes Assigmet Method for Solvig Travelig Salesma Problem Hadi Basirzadeh Departmet of Mathematics, Shahid Chamra Uiversity, Ahvaz, Ira Article history:
More informationCMSC Computer Architecture Lecture 12: Virtual Memory. Prof. Yanjing Li University of Chicago
CMSC 22200 Computer Architecture Lecture 12: Virtual Memory Prof. Yajig Li Uiversity of Chicago A System with Physical Memory Oly Examples: most Cray machies early PCs Memory early all embedded systems
More informationThe Magma Database file formats
The Magma Database file formats Adrew Gaylard, Bret Pikey, ad Mart-Mari Breedt Johaesburg, South Africa 15th May 2006 1 Summary Magma is a ope-source object database created by Chris Muller, of Kasas City,
More informationSorting in Linear Time. Data Structures and Algorithms Andrei Bulatov
Sortig i Liear Time Data Structures ad Algorithms Adrei Bulatov Algorithms Sortig i Liear Time 7-2 Compariso Sorts The oly test that all the algorithms we have cosidered so far is compariso The oly iformatio
More informationUH-MEM: Utility-Based Hybrid Memory Management. Yang Li, Saugata Ghose, Jongmoo Choi, Jin Sun, Hui Wang, Onur Mutlu
UH-MEM: Utility-Based Hybrid Memory Maagemet Yag Li, Saugata Ghose, Jogmoo Choi, Ji Su, Hui Wag, Our Mutlu 1 Executive Summary DRAM faces sigificat techology scalig difficulties Emergig memory techologies
More information3D Model Retrieval Method Based on Sample Prediction
20 Iteratioal Coferece o Computer Commuicatio ad Maagemet Proc.of CSIT vol.5 (20) (20) IACSIT Press, Sigapore 3D Model Retrieval Method Based o Sample Predictio Qigche Zhag, Ya Tag* School of Computer
More informationWhat are we going to learn? CSC Data Structures Analysis of Algorithms. Overview. Algorithm, and Inputs
What are we goig to lear? CSC316-003 Data Structures Aalysis of Algorithms Computer Sciece North Carolia State Uiversity Need to say that some algorithms are better tha others Criteria for evaluatio Structure
More informationMulti-Threading. Hyper-, Multi-, and Simultaneous Thread Execution
Multi-Threadig Hyper-, Multi-, ad Simultaeous Thread Executio 1 Performace To Date Icreasig processor performace Pipeliig. Brach predictio. Super-scalar executio. Out-of-order executio. Caches. Hyper-Threadig
More informationSession Initiated Protocol (SIP) and Message-based Load Balancing (MBLB)
F5 White Paper Sessio Iitiated Protocol (SIP) ad Message-based Load Balacig (MBLB) The ability to provide ew ad creative methods of commuicatios has esured a SIP presece i almost every orgaizatio. The
More informationPseudocode ( 1.1) Analysis of Algorithms. Primitive Operations. Pseudocode Details. Running Time ( 1.1) Estimating performance
Aalysis of Algorithms Iput Algorithm Output A algorithm is a step-by-step procedure for solvig a problem i a fiite amout of time. Pseudocode ( 1.1) High-level descriptio of a algorithm More structured
More informationL I N U X. Unit 6 S Y S T E M DHCP & DNS (BIND) A D M I N I S T R A T I O n DPW
it 6 HCP & (B) oa Warre HCP ervice yamically assigs a P address to requestig machies P addresses are leased scope of addresses ca be assiged or excluded from assigmet HCP servers do ot talk to each other
More informationNormal Distributions
Normal Distributios Stacey Hacock Look at these three differet data sets Each histogram is overlaid with a curve : A B C A) Weights (g) of ewly bor lab rat pups B) Mea aual temperatures ( F ) i A Arbor,
More information1&1 Next Level Hosting
1&1 Next Level Hostig Performace Level: Performace that grows with your requiremets Copyright 1&1 Iteret SE 2017 1ad1.com 2 1&1 NEXT LEVEL HOSTING 3 Fast page loadig ad short respose times play importat
More informationMaster Informatics Eng. 2017/18. A.J.Proença. Memory Hierarchy. (most slides are borrowed) AJProença, Advanced Architectures, MiEI, UMinho, 2017/18 1
Advaced Architectures Master Iformatics Eg. 2017/18 A.J.Proeça Memory Hierarchy (most slides are borrowed) AJProeça, Advaced Architectures, MiEI, UMiho, 2017/18 1 Itroductio Programmers wat ulimited amouts
More informationUNIT 4 Section 8 Estimating Population Parameters using Confidence Intervals
UNIT 4 Sectio 8 Estimatig Populatio Parameters usig Cofidece Itervals To make ifereces about a populatio that caot be surveyed etirely, sample statistics ca be take from a SRS of the populatio ad used
More informationPerformance Plus Software Parameter Definitions
Performace Plus+ Software Parameter Defiitios/ Performace Plus Software Parameter Defiitios Chapma Techical Note-TG-5 paramete.doc ev-0-03 Performace Plus+ Software Parameter Defiitios/2 Backgroud ad Defiitios
More informationAdministrative UNSUPERVISED LEARNING. Unsupervised learning. Supervised learning 11/25/13. Final project. No office hours today
Admiistrative Fial project No office hours today UNSUPERVISED LEARNING David Kauchak CS 451 Fall 2013 Supervised learig Usupervised learig label label 1 label 3 model/ predictor label 4 label 5 Supervised
More informationAlgorithms for Disk Covering Problems with the Most Points
Algorithms for Disk Coverig Problems with the Most Poits Bi Xiao Departmet of Computig Hog Kog Polytechic Uiversity Hug Hom, Kowloo, Hog Kog csbxiao@comp.polyu.edu.hk Qigfeg Zhuge, Yi He, Zili Shao, Edwi
More informationLower Bounds for Sorting
Liear Sortig Topics Covered: Lower Bouds for Sortig Coutig Sort Radix Sort Bucket Sort Lower Bouds for Sortig Compariso vs. o-compariso sortig Decisio tree model Worst case lower boud Compariso Sortig
More informationEM375 STATISTICS AND MEASUREMENT UNCERTAINTY LEAST SQUARES LINEAR REGRESSION ANALYSIS
EM375 STATISTICS AND MEASUREMENT UNCERTAINTY LEAST SQUARES LINEAR REGRESSION ANALYSIS I this uit of the course we ivestigate fittig a straight lie to measured (x, y) data pairs. The equatio we wat to fit
More informationArchitectural styles for software systems The client-server style
Architectural styles for software systems The cliet-server style Prof. Paolo Ciacarii Software Architecture CdL M Iformatica Uiversità di Bologa Ageda Cliet server style CS two tiers CS three tiers CS
More informationBGP with an Adaptive Minimal Route Advertisement Interval
with a Adaptive Miimal Route Advertisemet Iterval Nead Lasković ad Ljiljaa Trajković Simo Fraser Uiversity Vacouver, British Columbia, Caada {laskovi, ljilja}@cs.sfu.ca Abstract The duratio of the Miimal
More informationData Structures and Algorithms. Analysis of Algorithms
Data Structures ad Algorithms Aalysis of Algorithms Outlie Ruig time Pseudo-code Big-oh otatio Big-theta otatio Big-omega otatio Asymptotic algorithm aalysis Aalysis of Algorithms Iput Algorithm Output
More informationAdaptive and Lazy Segmentation Based Proxy Caching for Streaming Media Delivery
Adaptive ad Based Proxy Cachig for Streamig Media Delivery Sogqig Che Dept. of Computer Sciece College of William ad Mary Williamsburg, VA 23187 sqche@cs.wm.edu Susie Wee Mobile ad Media System Lab Hewlett-Packard
More informationBayesian approach to reliability modelling for a probability of failure on demand parameter
Bayesia approach to reliability modellig for a probability of failure o demad parameter BÖRCSÖK J., SCHAEFER S. Departmet of Computer Architecture ad System Programmig Uiversity Kassel, Wilhelmshöher Allee
More informationDATA MINING II - 1DL460
DATA MINING II - 1DL460 Sprig 2017 A secod course i data miig http://www.it.uu.se/edu/course/homepage/ifoutv2/vt17/ Kjell Orsbor Uppsala Database Laboratory Departmet of Iformatio Techology, Uppsala Uiversity,
More informationWeb Caching and Content Delivery
Web Caching and Content Delivery Caching for a Better Web Performance is a major concern in the Web Proxy caching is the most widely used method to improve Web performance Duplicate requests to the same
More informationLecture 5. Counting Sort / Radix Sort
Lecture 5. Coutig Sort / Radix Sort T. H. Corme, C. E. Leiserso ad R. L. Rivest Itroductio to Algorithms, 3rd Editio, MIT Press, 2009 Sugkyukwa Uiversity Hyuseug Choo choo@skku.edu Copyright 2000-2018
More informationAnalysis of Server Resource Consumption of Meteorological Satellite Application System Based on Contour Curve
Advaces i Computer, Sigals ad Systems (2018) 2: 19-25 Clausius Scietific Press, Caada Aalysis of Server Resource Cosumptio of Meteorological Satellite Applicatio System Based o Cotour Curve Xiagag Zhao
More information15-859E: Advanced Algorithms CMU, Spring 2015 Lecture #2: Randomized MST and MST Verification January 14, 2015
15-859E: Advaced Algorithms CMU, Sprig 2015 Lecture #2: Radomized MST ad MST Verificatio Jauary 14, 2015 Lecturer: Aupam Gupta Scribe: Yu Zhao 1 Prelimiaries I this lecture we are talkig about two cotets:
More informationOperating System Concepts. Operating System Concepts
Chapter 4: Mass-Storage Systems Logical Disk Structure Logical Disk Structure Disk Schedulig Disk Maagemet RAID Structure Disk drives are addressed as large -dimesioal arrays of logical blocks, where the
More informationA PREDICTION MODEL FOR USER S SHARE ANALYSIS IN DUAL- SIM ENVIRONMENT
GSJ: Computer Sciece ad Telecommuicatios 03 No.3(39) ISSN 5-3 A PRDICTION MODL FOR USR S SHAR ANALYSIS IN DUAL- SIM NVIRONMNT Thakur Sajay, Jai Parag Orietal Uiversity, Idore, Idia sajaymca00@yahoo.com
More informationECE4050 Data Structures and Algorithms. Lecture 6: Searching
ECE4050 Data Structures ad Algorithms Lecture 6: Searchig 1 Search Give: Distict keys k 1, k 2,, k ad collectio L of records of the form (k 1, I 1 ), (k 2, I 2 ),, (k, I ) where I j is the iformatio associated
More informationDCMIX: Generating Mixed Workloads for the Cloud Data Center
DCMIX: Geeratig Mixed Workloads for the Cloud Data Ceter XigWag Xiog, Lei Wag, WaLig Gao, Rui Re, Ke Liu, Che Zheg, Yu We, YiLiag Istitute of Computig Techology, Chiese Academy of Scieces Bech 2018, Seattle,
More informationThe Value of Peering
The Value of Peerig ISP/IXP Workshops These materials are licesed uder the Creative Commos Attributio-NoCommercial 4.0 Iteratioal licese (http://creativecommos.org/liceses/by-c/4.0/) Last updated 25 th
More informationThe Penta-S: A Scalable Crossbar Network for Distributed Shared Memory Multiprocessor Systems
The Peta-S: A Scalable Crossbar Network for Distributed Shared Memory Multiprocessor Systems Abdulkarim Ayyad Departmet of Computer Egieerig, Al-Quds Uiversity, Jerusalem, P.O. Box 20002 Tel: 02-2797024,
More informationAn Improved Shuffled Frog-Leaping Algorithm for Knapsack Problem
A Improved Shuffled Frog-Leapig Algorithm for Kapsack Problem Zhoufag Li, Ya Zhou, ad Peg Cheg School of Iformatio Sciece ad Egieerig Hea Uiversity of Techology ZhegZhou, Chia lzhf1978@126.com Abstract.
More informationBig-O Analysis. Asymptotics
Big-O Aalysis 1 Defiitio: Suppose that f() ad g() are oegative fuctios of. The we say that f() is O(g()) provided that there are costats C > 0 ad N > 0 such that for all > N, f() Cg(). Big-O expresses
More informationOCR Statistics 1. Working with data. Section 3: Measures of spread
Notes ad Eamples OCR Statistics 1 Workig with data Sectio 3: Measures of spread Just as there are several differet measures of cetral tedec (averages), there are a variet of statistical measures of spread.
More informationSwitching Hardware. Spring 2018 CS 438 Staff, University of Illinois 1
Switchig Hardware Sprig 208 CS 438 Staff, Uiversity of Illiois Where are we? Uderstad Differet ways to move through a etwork (forwardig) Read sigs at each switch (datagram) Follow a kow path (virtual circuit)
More informationIMP: Superposer Integrated Morphometrics Package Superposition Tool
IMP: Superposer Itegrated Morphometrics Package Superpositio Tool Programmig by: David Lieber ( 03) Caisius College 200 Mai St. Buffalo, NY 4208 Cocept by: H. David Sheets, Dept. of Physics, Caisius College
More information9.1. Sequences and Series. Sequences. What you should learn. Why you should learn it. Definition of Sequence
_9.qxd // : AM Page Chapter 9 Sequeces, Series, ad Probability 9. Sequeces ad Series What you should lear Use sequece otatio to write the terms of sequeces. Use factorial otatio. Use summatio otatio to
More informationOur second algorithm. Comp 135 Machine Learning Computer Science Tufts University. Decision Trees. Decision Trees. Decision Trees.
Comp 135 Machie Learig Computer Sciece Tufts Uiversity Fall 2017 Roi Khardo Some of these slides were adapted from previous slides by Carla Brodley Our secod algorithm Let s look at a simple dataset for
More informationMath Section 2.2 Polynomial Functions
Math 1330 - Sectio. Polyomial Fuctios Our objectives i workig with polyomial fuctios will be, first, to gather iformatio about the graph of the fuctio ad, secod, to use that iformatio to geerate a reasoably
More informationCS2410 Computer Architecture. Flynn s Taxonomy
CS2410 Computer Architecture Dept. of Computer Sciece Uiversity of Pittsburgh http://www.cs.pitt.edu/~melhem/courses/2410p/idex.html 1 Fly s Taxoomy SISD Sigle istructio stream Sigle data stream (SIMD)
More informationA New Morphological 3D Shape Decomposition: Grayscale Interframe Interpolation Method
A ew Morphological 3D Shape Decompositio: Grayscale Iterframe Iterpolatio Method D.. Vizireau Politehica Uiversity Bucharest, Romaia ae@comm.pub.ro R. M. Udrea Politehica Uiversity Bucharest, Romaia mihea@comm.pub.ro
More informationImproving Template Based Spike Detection
Improvig Template Based Spike Detectio Kirk Smith, Member - IEEE Portlad State Uiversity petra@ee.pdx.edu Abstract Template matchig algorithms like SSE, Covolutio ad Maximum Likelihood are well kow for
More informationRandom Graphs and Complex Networks T
Radom Graphs ad Complex Networks T-79.7003 Charalampos E. Tsourakakis Aalto Uiversity Lecture 3 7 September 013 Aoucemet Homework 1 is out, due i two weeks from ow. Exercises: Probabilistic iequalities
More informationArecent solicitation from the National Science Foundation
Cliet-Server Computig WWW Traffic Reductio ad Load Balacig through Server-Based Cachig Azer Bestavros Bosto Uiversity This cachig protocol exploits the geographic ad temporal locality of referece exhibited
More informationTutorial on Packet Time Metrics
Power Matters. Tutorial o Packet Time Metrics Lee Cosart lee.cosart@microsemi.com ITS 204 204 Microsemi Corporatio. COMPANY POPIETAY Itroductio requecy trasport Oe-way: forward & reverse packet streams
More informationBGP Attributes and Path Selection. ISP Training Workshops
BGP Attributes ad Path Selectio ISP Traiig Workshops 1 BGP Attributes The tools available for the job 2 What Is a Attribute?... Next Hop AS Path MED...... p Part of a BGP Update p Describes the characteristics
More informationSAMPLE VERSUS POPULATION. Population - consists of all possible measurements that can be made on a particular item or procedure.
SAMPLE VERSUS POPULATION Populatio - cosists of all possible measuremets that ca be made o a particular item or procedure. Ofte a populatio has a ifiite umber of data elemets Geerally expese to determie
More informationAnalysis Metrics. Intro to Algorithm Analysis. Slides. 12. Alg Analysis. 12. Alg Analysis
Itro to Algorithm Aalysis Aalysis Metrics Slides. Table of Cotets. Aalysis Metrics 3. Exact Aalysis Rules 4. Simple Summatio 5. Summatio Formulas 6. Order of Magitude 7. Big-O otatio 8. Big-O Theorems
More informationThe golden search method: Question 1
1. Golde Sectio Search for the Mode of a Fuctio The golde search method: Questio 1 Suppose the last pair of poits at which we have a fuctio evaluatio is x(), y(). The accordig to the method, If f(x())
More informationCS61C : Machine Structures
CS 61C L24 VM II (1) ist.eecs.berkele.edu/~cs61c/su5 CS61C : Machie Structures Lecture #24: VM II Address Mappig: Virtual Address: VPN offset 25-8-2 Ad Carle idex ito page table located i phsical memor
More informationEvaluation of Support Vector Machine Kernels for Detecting Network Anomalies
Evaluatio of Support Vector Machie Kerels for Detectig Network Aomalies Prera Batta, Maider Sigh, Zhida Li, Qigye Dig, ad Ljiljaa Trajković Commuicatio Networks Laboratory http://www.esc.sfu.ca/~ljilja/cl/
More informationA Study on the Performance of Cholesky-Factorization using MPI
A Study o the Performace of Cholesky-Factorizatio usig MPI Ha S. Kim Scott B. Bade Departmet of Computer Sciece ad Egieerig Uiversity of Califoria Sa Diego {hskim, bade}@cs.ucsd.edu Abstract Cholesky-factorizatio
More informationUNIVERSITY OF MORATUWA
UNIVERSITY OF MORATUWA FACULTY OF ENGINEERING DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING B.Sc. Egieerig 2014 Itake Semester 2 Examiatio CS2052 COMPUTER ARCHITECTURE Time allowed: 2 Hours Jauary 2016
More informationComputer Architecture ELEC3441
CPU-Memory Bottleeck Computer Architecture ELEC44 CPU Memory Lecture 8 Cache Dr. Hayde Kwok-Hay So Departmet of Electrical ad Electroic Egieerig Performace of high-speed computers is usually limited by
More informationWavelet Transform. CSE 490 G Introduction to Data Compression Winter Wavelet Transformed Barbara (Enhanced) Wavelet Transformed Barbara (Actual)
Wavelet Trasform CSE 49 G Itroductio to Data Compressio Witer 6 Wavelet Trasform Codig PACW Wavelet Trasform A family of atios that filters the data ito low resolutio data plus detail data high pass filter
More informationDescriptive Statistics Summary Lists
Chapter 209 Descriptive Statistics Summary Lists Itroductio This procedure is used to summarize cotiuous data. Large volumes of such data may be easily summarized i statistical lists of meas, couts, stadard
More informationProtected points in ordered trees
Applied Mathematics Letters 008 56 50 www.elsevier.com/locate/aml Protected poits i ordered trees Gi-Sag Cheo a, Louis W. Shapiro b, a Departmet of Mathematics, Sugkyukwa Uiversity, Suwo 440-746, Republic
More informationData diverse software fault tolerance techniques
Data diverse software fault tolerace techiques Complemets desig diversity by compesatig for desig diversity s s limitatios Ivolves obtaiig a related set of poits i the program data space, executig the
More informationRealistic Storage of Pending Requests in Content-Centric Network Routers
Realistic Storage of Pedig Requests i Cotet-Cetric Network Routers Wei You, Bertrad Mathieu, Patrick Truog, Jea-Fraçois Peltier Orage Labs Laio, Frace {wei.you, bertrad2.mathieu, patrick.truog, jeafracois.peltier}@orage.com
More informationCourse Site: Copyright 2012, Elsevier Inc. All rights reserved.
Course Site: http://cc.sjtu.edu.c/g2s/site/aca.html 1 Computer Architecture A Quatitative Approach, Fifth Editio Chapter 2 Memory Hierarchy Desig 2 Outlie Memory Hierarchy Cache Desig Basic Cache Optimizatios
More informationTraffic Models and QoS
Traffic Models ad QoS TELE4642: Week9 Ackowledgemet: Some slides are adapted from Computer Networkig: A Top Dow Approach Featurig the Iteret, 2 d editio, J.F Kurose ad K.W. Ross All Rights Reserved, copyright
More informationprerequisites: 6.046, 6.041/2, ability to do proofs Randomized algorithms: make random choices during run. Main benefits:
Itro Admiistrivia. Sigup sheet. prerequisites: 6.046, 6.041/2, ability to do proofs homework weekly (first ext week) collaboratio idepedet homeworks gradig requiremet term project books. questio: scribig?
More informationChapter 3 Classification of FFT Processor Algorithms
Chapter Classificatio of FFT Processor Algorithms The computatioal complexity of the Discrete Fourier trasform (DFT) is very high. It requires () 2 complex multiplicatios ad () complex additios [5]. As
More informationData Analysis. Concepts and Techniques. Chapter 2. Chapter 2: Getting to Know Your Data. Data Objects and Attribute Types
Data Aalysis Cocepts ad Techiques Chapter 2 1 Chapter 2: Gettig to Kow Your Data Data Objects ad Attribute Types Basic Statistical Descriptios of Data Data Visualizatio Measurig Data Similarity ad Dissimilarity
More informationFundamentals of Media Processing. Shin'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dinh Le
Fudametals of Media Processig Shi'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dih Le Today's topics Noparametric Methods Parze Widow k-nearest Neighbor Estimatio Clusterig Techiques k-meas Agglomerative Hierarchical
More informationLecture Notes 6 Introduction to algorithm analysis CSS 501 Data Structures and Object-Oriented Programming
Lecture Notes 6 Itroductio to algorithm aalysis CSS 501 Data Structures ad Object-Orieted Programmig Readig for this lecture: Carrao, Chapter 10 To be covered i this lecture: Itroductio to algorithm aalysis
More informationAdaptive Resource Allocation for Electric Environmental Pollution through the Control Network
Available olie at www.sciecedirect.com Eergy Procedia 6 (202) 60 64 202 Iteratioal Coferece o Future Eergy, Eviromet, ad Materials Adaptive Resource Allocatio for Electric Evirometal Pollutio through the
More informationLoad balanced Parallel Prime Number Generator with Sieve of Eratosthenes on Cluster Computers *
Load balaced Parallel Prime umber Geerator with Sieve of Eratosthees o luster omputers * Soowook Hwag*, Kyusik hug**, ad Dogseug Kim* *Departmet of Electrical Egieerig Korea Uiversity Seoul, -, Rep. of
More informationNON-LINEAR MODELLING OF A GEOTHERMAL STEAM PIPE
14thNew Zealad Workshop 1992 NON-LNEAR MODELLNG OF A GEOTHERMAL STEAM PPE Y. Huag ad D. H. Freesto Geothermal stitute, Uiversity of Aucklad SUMMARY Recet work o developig a o-liear model for a geothermal
More information. Written in factored form it is easy to see that the roots are 2, 2, i,
CMPS A Itroductio to Programmig Programmig Assigmet 4 I this assigmet you will write a java program that determies the real roots of a polyomial that lie withi a specified rage. Recall that the roots (or
More informationRunning Time ( 3.1) Analysis of Algorithms. Experimental Studies. Limitations of Experiments
Ruig Time ( 3.1) Aalysis of Algorithms Iput Algorithm Output A algorithm is a step- by- step procedure for solvig a problem i a fiite amout of time. Most algorithms trasform iput objects ito output objects.
More informationAnalysis of Algorithms
Aalysis of Algorithms Iput Algorithm Output A algorithm is a step-by-step procedure for solvig a problem i a fiite amout of time. Ruig Time Most algorithms trasform iput objects ito output objects. The
More informationAnnouncements. Reading. Project #4 is on the web. Homework #1. Midterm #2. Chapter 4 ( ) Note policy about project #3 missing components
Aoucemets Readig Chapter 4 (4.1-4.2) Project #4 is o the web ote policy about project #3 missig compoets Homework #1 Due 11/6/01 Chapter 6: 4, 12, 24, 37 Midterm #2 11/8/01 i class 1 Project #4 otes IPv6Iit,
More informationHash Tables. Presentation for use with the textbook Algorithm Design and Applications, by M. T. Goodrich and R. Tamassia, Wiley, 2015.
Presetatio for use with the textbook Algorithm Desig ad Applicatios, by M. T. Goodrich ad R. Tamassia, Wiley, 2015 Hash Tables xkcd. http://xkcd.com/221/. Radom Number. Used with permissio uder Creative
More informationA General Framework for Accurate Statistical Timing Analysis Considering Correlations
A Geeral Framework for Accurate Statistical Timig Aalysis Cosiderig Correlatios 7.4 Vishal Khadelwal Departmet of ECE Uiversity of Marylad-College Park vishalk@glue.umd.edu Akur Srivastava Departmet of
More informationEvaluation scheme for Tracking in AMI
A M I C o m m u i c a t i o A U G M E N T E D M U L T I - P A R T Y I N T E R A C T I O N http://www.amiproject.org/ Evaluatio scheme for Trackig i AMI S. Schreiber a D. Gatica-Perez b AMI WP4 Trackig:
More informationEnhancing Cloud Computing Scheduling based on Queuing Models
Ehacig Cloud Computig Schedulig based o Queuig Models Mohamed Eisa Computer Sciece Departmet, Port Said Uiversity, 42526 Port Said, Egypt E. I. Esedimy Computer Sciece Departmet, Masoura Uiversity, Masoura,
More informationDesigning a learning system
CS 75 Itro to Machie Learig Lecture Desigig a learig system Milos Hauskrecht milos@pitt.edu 539 Seott Square, -5 people.cs.pitt.edu/~milos/courses/cs75/ Admiistrivia No homework assigmet this week Please
More informationChapter 10. Defining Classes. Copyright 2015 Pearson Education, Ltd.. All rights reserved.
Chapter 10 Defiig Classes Copyright 2015 Pearso Educatio, Ltd.. All rights reserved. Overview 10.1 Structures 10.2 Classes 10.3 Abstract Data Types 10.4 Itroductio to Iheritace Copyright 2015 Pearso Educatio,
More informationCS200: Hash Tables. Prichard Ch CS200 - Hash Tables 1
CS200: Hash Tables Prichard Ch. 13.2 CS200 - Hash Tables 1 Table Implemetatios: average cases Search Add Remove Sorted array-based Usorted array-based Balaced Search Trees O(log ) O() O() O() O(1) O()
More informationQuality of Service. Spring 2018 CS 438 Staff - University of Illinois 1
Quality of Service Sprig 2018 CS 438 Staff - Uiversity of Illiois 1 Quality of Service How good are late data ad lowthroughput chaels? It depeds o the applicatio. Do you care if... Your e-mail takes 1/2
More informationDesigning a learning system
CS 75 Machie Learig Lecture Desigig a learig system Milos Hauskrecht milos@cs.pitt.edu 539 Seott Square, x-5 people.cs.pitt.edu/~milos/courses/cs75/ Admiistrivia No homework assigmet this week Please try
More informationn Based on unrealistic growth forecast n Overcapacity: Fiber 5x100 in three years n Wireless: Expensive spectrum licenses n Fibers
EECS228a Research Topics Jea Walrad www.eecs.berkeley.edu/~wlr of Networks Walrad 52 of Networks Pricig of Services Competitio of Users Competitio of Providers Suggested Readigs: http://www.bgsu.edu/departmets/tcom/aota.htm
More informationOn the Throughput Capacity of Information-Centric Networks
O the Throughput Capacity of Iformatio-Cetric Networks Bita Azimdoost, Cedric estphal, ad Hamid R. Sadjadpour Departmet of Electrical Egieerig ad Computer Egieerig Uiversity of Califoria Sata Cruz, Sata
More informationFPGA IMPLEMENTATION OF BASE-N LOGARITHM. Salvador E. Tropea
FPGA IMPLEMENTATION OF BASE-N LOGARITHM Salvador E. Tropea Electróica e Iformática Istituto Nacioal de Tecología Idustrial Bueos Aires, Argetia email: salvador@iti.gov.ar ABSTRACT I this work, we preset
More informationReliable Transmission. Spring 2018 CS 438 Staff - University of Illinois 1
Reliable Trasmissio Sprig 2018 CS 438 Staff - Uiversity of Illiois 1 Reliable Trasmissio Hello! My computer s ame is Alice. Alice Bob Hello! Alice. Sprig 2018 CS 438 Staff - Uiversity of Illiois 2 Reliable
More informationCh 9.3 Geometric Sequences and Series Lessons
Ch 9.3 Geometric Sequeces ad Series Lessos SKILLS OBJECTIVES Recogize a geometric sequece. Fid the geeral, th term of a geometric sequece. Evaluate a fiite geometric series. Evaluate a ifiite geometric
More informationService Oriented Enterprise Architecture and Service Oriented Enterprise
Approved for Public Release Distributio Ulimited Case Number: 09-2786 The 23 rd Ope Group Eterprise Practitioers Coferece Service Orieted Eterprise ad Service Orieted Eterprise Ya Zhao, PhD Pricipal, MITRE
More informationRunning Time. Analysis of Algorithms. Experimental Studies. Limitations of Experiments
Ruig Time Aalysis of Algorithms Iput Algorithm Output A algorithm is a step-by-step procedure for solvig a problem i a fiite amout of time. Most algorithms trasform iput objects ito output objects. The
More information