Presentation is loading. Please wait.

Presentation is loading. Please wait.

How Fair is your Queue Hanoch Levy

Similar presentations


Presentation on theme: "How Fair is your Queue Hanoch Levy"— Presentation transcript:

1 How Fair is your Queue Hanoch Levy
School of Computer Science, Tel Aviv University Jointly with Benjamin Avi-Itzhak, RUTGERS University David Raz, Tel-Aviv University Alternative name: “Satellite-linked Web Caches” March, 2004 1/16/2019 Fairness in Queues, H. Levy

2 To provide FAIRNESS in waiting/service
Why QUEUES? “Not Fair!!!” “This is more Fair…” To provide FAIRNESS in waiting/service Sophisticated queues to address fairness issues 1/16/2019 Fairness in Queues, H. Levy

3 Queueing Theory, queues and fairness
Queueing theory: Decades of very deep and elaborate research Queueing structures / policies, distributions Focus on delay of individual: Moments/ distribution / optimal operations, many more! Practical Applications: Efficient control / operation of: Bank, computer system, web server, telecom Fairness in queues: Many statements: “this is fair”, “that system is unfair”. Very little analysis (job fairness): Larson (1988), Palm (1953), Mann (1969), Whitt (1984): Discuss justice related / overtaking Morris & Wang (85). Avi-Itzhak & Levy (96) Wierman & Harchol-Balter (Sigmetrics 2003):  We don’t know how to quantify queue fairness! 1/16/2019 Fairness in Queues, H. Levy

4 How Important is fairness in queues?
FAIRNESS INHERENT/CRUCIAL part of queues Recent studies, Rafaeli et. al. [2003] (experimental psychology): Experiments on humans in multi-queue and single queue Fairness in queue is very important to people Perhaps even more than delay itself! WFQ: 10’s of papers – fairness among jobs whose duration is O(1) microsecond Economic value of queues of humans: O(1%) of GNP? 1/16/2019 Fairness in Queues, H. Levy

5 “So – What is the problem? “
“Take social-science/economics utility-fairness measures and apply to queues” HOW??? What is the “PIE”? A “moving target”… The problem in a nutshell: Short vs. Long Long Short The difficulty (A): Whom to compare a customer against? The difficulty (B): Size vs. seniority dilemma 1/16/2019 Fairness in Queues, H. Levy

6 Motivation / Questions
Should Long be served ahead of Short? Short(S) Long (L) Is it fair? How fair (how much fair) is it to serve Short Ahead of Long? Quantify/ Measure Fairness in Waiting lines! 1/16/2019 Fairness in Queues, H. Levy

7 Motivation / Questions (2): Do You like your supermarket?
How fair is a scheduling policy? This? E.g: FIFO LIFO Multiple Queue (Multi server) Single queue (multi server) Queues by job size? Or This? 1/16/2019 Fairness in Queues, H. Levy

8 Motivation / Questions (3): and your airport?
Multiple Queue (Multi server) Single queue (multi server) Smith & Whitt (81), Larson (87), Rothkopf & Rech (87), Wolff (77, 87, 88) 1/16/2019 Fairness in Queues, H. Levy

9 Applications (computer world)
Internet revolution: Service shift to computer systems. Responsibility of control: shifts to computer programmer/operator Examples: Call centers: Web services: How should I operate my web-server? FIFO? LIFO? Priorities? Call center Web server 1/16/2019 Fairness in Queues, H. Levy

10 Fairness in Queues, H. Levy
Related Work(1) Avi-Itzhak & Levy (96): Axiomatic approach Departure point+emphasis: Seniority (Order of service) If service times are identical  variance of waiting time measures fairness Extend to service times When is it good? 1/16/2019 Fairness in Queues, H. Levy

11 Fairness in Queues, H. Levy
Recent Related Work Wierman & Harchol-Balter (Sigmetrics 2003): Propose a Fairness Criterion Class-based approach: For job of size x compute E[T(x)/x] If this is bounded by 1/(1-rho) for all x  FAIR. Results: Analyze the classification for a large variety of policies. FIFO (FCFS) – is “Always UNFAIR” LIFO (preemptive) – is FAIR. When is it good? 1/16/2019 Fairness in Queues, H. Levy

12 Conflicting (disturbing) Views
Policy Ordinary person Queueing theory (WHB 03) FIFO LIFO (preemptive) Fair (Fairest?) Unfair Unfair (most?) Fair 1/16/2019 Fairness in Queues, H. Levy

13 Basic Requirements of a Queue Fairness measure
Aim for standard:  Have a consistent view/intuition Deal with individual, scenario, and system Account for both seniority and service requirement: Seniority: Service times are identical: Fairness is a function of seniority FIFO most fair / LIFO most unfair Senior ahead of junior is more fair Service requirement: Arrival times are identical Fairness is a function of service requirement Short ahead of long is more fair Yields to analysis 1/16/2019 Fairness in Queues, H. Levy

14 RAQFM: A Resource Allocation Queueing Fairness Measure
Aims at meeting these requirements In particular: Long vs. Short Seniority vs. service times S L 1/16/2019 Fairness in Queues, H. Levy

15 Approach: individual discrimination
To whom should a job be compared? (moving target!) Compare to the “public”. Focus on server resources (aim at equal division) Weigh the warranted service with the granted service Equal Share philosophy (“axiom”): at t: Warranted service of Granted service of Individual discrimination of (net service) 1/16/2019 Fairness in Queues, H. Levy

16 Basic Properties of RAQFM
Resource allocation is proper = zero-sum discrimination (work conserving, non idling) Individual discrimination: We are trying to use SRM to analyze the CSDS System measure of discrimination: aggregate statistics of Eliminate: Expected discrimination Reasonable: distance from mean, Var(D), E[|D-E[D]|]. Measure of Unfairness 1/16/2019 Fairness in Queues, H. Levy

17 Property 1: Processor Sharing: Ultimate Fairness
Single server queue Processor sharing service policy (Kleinrock (64), (67), Coffman, Muntz & Trotter (70)) Theorem 1: Under PS  Var[D] = 0  PS is the most fair policy!! We are trying to use SRM to analyze the CSDS 1/16/2019 Fairness in Queues, H. Levy

18 Property 2: SENIORITY (identical service times)
Single server queue Service times are all identical Arrival times are arbitrary Theorem 2: Serving by order of seniority (FIFO) is most fair Serving in reverse order of seniority (LIFO) is most unfair Pushing a junior ahead of senior reduces fairness We are trying to use SRM to analyze the CSDS 1/16/2019 Fairness in Queues, H. Levy

19 Property 3: Service Time (identical arrival times)
Single server queue Arrival times are all identical Service times are arbitrary S L Theorem 3: Serving shortest job first (SJF) is most fair Serving Longest job first (LJF) is most unfair Pushing a large job ahead of small job reduces fairness We are trying to use SRM to analyze the CSDS 1/16/2019 Fairness in Queues, H. Levy

20 Property 4: More advanced The Long vs. Short case
Single server queue Long and short arrive at different times S L We are trying to use SRM to analyze the CSDS Fairness of two orders depends on relative seniority and relative service times. 1/16/2019 Fairness in Queues, H. Levy

21 Property 5: Bounds How bad (good) can it be?
Bound on individual discrimination Use for scale of reference / sanity check How good: (service time) How bad: (waiting time) We are trying to use SRM to analyze the CSDS Bound on system discrimination: How bad: 1/16/2019 Fairness in Queues, H. Levy

22 Property 6: Locality of Comparison
Measure can be evaluated by comparing all customers (across busy periods) Unique to RAQFM within a large function family. Important for fairness: One should compare only relevant customers (within busy periods). We are trying to use SRM to analyze the CSDS 1/16/2019 Fairness in Queues, H. Levy

23 Property 7: Discrimination Monotonic in Service time
THM: For an arbitrary system, if service decision is independent of service time, then: Discrimination monotonically increases with service time (deterministic) Larger customers get preferential service Discrimination monotonically increases with service time distribution  justification to the prioritization of short jobs! We are trying to use SRM to analyze the CSDS 1/16/2019 Fairness in Queues, H. Levy

24 Fairness in Queues, H. Levy
RAQFM is Analyzable RAQFM yields to analysis via standard queueing theory techniques Can compute E[D| x] (class discrimination) Var[D] (system unfairness) Conducted analysis for M/M/1 type: Variety of service orders (FIFO, LIFO, ROS, more…). Conducted analysis for Multi-queue / multi-class systems. We are trying to use SRM to analyze the CSDS 1/16/2019 Fairness in Queues, H. Levy

25 Individual discrimination under FIFO: M/M/1 ( conditioned on # customers found ahead)
Utilization Indifferent Time to smile Empty super-market Friday noon Time to cry Traffic jam at 4AM Discrimination as a function of # customers found at queue 1/16/2019 Fairness in Queues, H. Levy

26 System Unfairness: Compare operation policies
PS: Absolute Fairness! LIFO: Severe seniority discrimination Empty system: everyone is alone FIFO: no seniority discrimination System Unfairness as a function of system load 1/16/2019 Fairness in Queues, H. Levy

27 How other measures relate: Bridging the gap
Wierman & Harchol-Balter (2003) FIFO (FCFS) – is UNFAIR LIFO (preemptive) – is FAIR LIFO (preemptive) FIFO Queueing theory (WHB 2003) Ordinary person View Policy Unfair  Fair Unfair Seniority Service times 1/16/2019 Fairness in Queues, H. Levy

28 RAQFM: account for all factors - bridge the gap
Seniority + service time differences play role (MOST CASES!)  RAQFM agrees with ordinary person FIFO: LIFO: Service time differences very radical (A few cases)  RAQFM agrees with Wierman & Harchol Balter 1/16/2019 Fairness in Queues, H. Levy

29 Fairness in Queues, H. Levy
Comparison of Methods [AL96]: (SENIORITY) Easy to compute Order fairness: When the issue is ORDER [WHB03]: (SERVICE TIMES) When jobs do not see each other / do not care of each other. RAQFM: (SENIORITY & SERVICE TIMES) Somewhat harder to compute When issue is waiting time. Also for ORDER fairness. 1/16/2019 Fairness in Queues, H. Levy

30 Fairness in Queues, H. Levy
Summary Fairness is fundamental for queueing systems No agreed upon measure exists RAQFM is a queueing fairness measure that: Has a consistent view Deals with individuals, scenario, and system Accounts for both seniority and service requirement Admits logically to special cases: Service times are identical: Senior ahead of junior is more fair Arrival times are identical Short ahead of long is more fair Yields to analysis We analyzed a large variety of queueing systems Much more work is needed 1/16/2019 Fairness in Queues, H. Levy

31 Fairness in Queues, H. Levy
Closing Words Thank you If you have applications at which fairness is relevant – we will be glad to hear. Whenever you “enjoy” the queues of your supermarket / bank / airport / … give us a call… 1/16/2019 Fairness in Queues, H. Levy


Download ppt "How Fair is your Queue Hanoch Levy"

Similar presentations


Ads by Google