, Volume 81, Issue 9, pp 3391–3421 | Cite as

Primal–Dual and Dual-Fitting Analysis of Online Scheduling Algorithms for Generalized Flow-Time Problems

  • Spyros Angelopoulos
  • Giorgio LucarelliEmail author
  • Nguyen Kim Thang 


We study online scheduling problems on a single processor that can be viewed as extensions of the well-studied problem of minimizing total weighted flow time. In particular, we provide a framework of analysis that is derived by duality properties, does not rely on potential functions and is applicable to a variety of scheduling problems. A key ingredient in our approach is bypassing the need for “black-box” rounding of fractional solutions, which yields improved competitive ratios. We begin with an interpretation of Highest-Density-First (HDF) as a primal–dual algorithm, and a corresponding proof that HDF is optimal for total fractional weighted flow time (and thus scalable for the integral objective). Building upon the salient ideas of the proof, we show how to apply and extend this analysis to the more general problem of minimizing \(\sum _j w_j g(F_j)\), where \(w_j\) is the job weight, \(F_j\) is the flow time and g is a non-decreasing cost function. Among other results, we present improved competitive ratios for the setting in which g is a concave function, and the setting of same-density jobs but general cost functions. We further apply our framework of analysis to online weighted completion time with general cost functions as well as scheduling under polyhedral constraints.


Online algorithms Primal–dual Dual-fitting Generalized flow-time Scheduling 



We would like to thank an anonymous referee for pointing out an error in an earlier version of this paper. Spyros Angelopoulos is supported by project ANR-11-BS02-0015 “New Techniques in Online Computation (NeTOC)”. Giorgio Lucarelli is supported by the ANR project Moebus (Grant No. ANR-13-INFR-0001). Nguyen Kim Thang supported by the FMJH program Gaspard Monge in optimization and operations research and by EDF.


  1. 1.
    Anand, S., Garg, N., Kumar, A.: Resource augmentation for weighted flow-time explained by dual fitting. In: SODA, pp. 1228–1241 (2012)Google Scholar
  2. 2.
    Antoniadis, A., Barcelo, N., Consuegra, M., Kling, P., Nugent, M., Pruhs, K., Scquizzato, M.: Efficient computation of optimal energy and fractional weighted flow trade-off schedules. In: STACS, vol. 25 of LIPIcs, pp. 63–74 (2014)Google Scholar
  3. 3.
    Bansal, N., Chan, H.-L.: Weighted flow time does not admit \(o(1)\)-competitive algorithms. In: SODA, pp. 1238–1244 (2009)Google Scholar
  4. 4.
    Bansal, N., Pruhs, K.: Server scheduling in the weighted \(\ell _p\) norm. In: LATIN, vol. 2976 of LNCS, pp. 434–443 (2004)Google Scholar
  5. 5.
    Bansal, N., Pruhs, K.: The geometry of scheduling. In: FOCS, pp. 407–414 (2010)Google Scholar
  6. 6.
    Becchetti, L., Leonardi, S., Marchetti-Spaccamela, A., Pruhs, K.: Online weighted flow time and deadline scheduling. J. Discrete Algorithms 4(3), 339–352 (2006)MathSciNetCrossRefzbMATHGoogle Scholar
  7. 7.
    Devanur, N.R., Huang, Z.: Primal dual gives almost optimal energy efficient online algorithms. In: SODA, pp. 1123–1140 (2014)Google Scholar
  8. 8.
    Edmonds, J.: Scheduling in the dark. Theor. Comput. Sci. 235(1), 109–141 (2000)MathSciNetCrossRefzbMATHGoogle Scholar
  9. 9.
    Edmonds, J., Pruhs, K.: Scalably scheduling processes with arbitrary speedup curves. ACM Trans. Algorithms 8(3), 28 (2012)MathSciNetCrossRefzbMATHGoogle Scholar
  10. 10.
    Fox, K., Im, S., Kulkarni, J., Moseley, B.: Online non-clairvoyant scheduling to simultaneously minimize all convex functions. In: APPROX-RANDOM, vol.8096 of LNCS, pp. 142–157 (2013)Google Scholar
  11. 11.
    Gupta, A., Im, S., Krishnaswamy, R., Moseley, B., Pruhs, K.: Scheduling heterogeneous processors isn’t as easy as you think. In: SODA, pp. 1242–1253 (2012)Google Scholar
  12. 12.
    Gupta, A., Krishnaswamy, R., Pruhs, K.: Nonclairvoyantly scheduling power-heterogeneous processors. In: Green Computing Conference, pp. 165–173 (2010)Google Scholar
  13. 13.
    Gupta, A., Krishnaswamy, R., Pruhs, K.: Online primal–dual for non-linear optimization with applications to speed scaling. In: WAOA, vol. 7846 of LNCS, pp. 173–186 (2012)Google Scholar
  14. 14.
    Höhn, W., Mestre, J., Wiese, A.: How unsplittable-flow-covering helps scheduling with job-dependent cost functions. In: ICALP, vol. 8572 of LNCS, pp. 625–636 (2014)Google Scholar
  15. 15.
    Im, S., Kulkarni, J., Munagala, K.: Competitive algorithms from competitive equilibria: Non-clairvoyant scheduling under polyhedral constraints. In: STOC, pp. 313–322 (2014)Google Scholar
  16. 16.
    Im, S., Kulkarni, J., Munagala, K., Pruhs, K.: Selfishmigrate: a scalable algorithm for non-clairvoyantly scheduling heterogeneous processors. In: FOCS, pp. 531–540 (2014)Google Scholar
  17. 17.
    Im, S., Moseley, B., Pruhs, K.: A tutorial on amortized local competitiveness in online scheduling. SIGACT News 42(2), 83–97 (2011)CrossRefGoogle Scholar
  18. 18.
    Im, S., Moseley, B., Pruhs, K.: Online scheduling with general cost functions. SIAM J. Comput. 43(1), 126–143 (2014)MathSciNetCrossRefzbMATHGoogle Scholar
  19. 19.
    Jain, K., Mahdian, M., Markakis, E., Saberi, A., Vazirani, V.V.: Greedy facility location algorithms analyzed using dual fitting with factor-revealing LP. JACM 50(6), 795–824 (2003)MathSciNetCrossRefzbMATHGoogle Scholar
  20. 20.
    Kalyanasundaram, B., Pruhs, K.: Speed is as powerful as clairvoyance. JACM 47(4), 617–643 (2000)MathSciNetCrossRefzbMATHGoogle Scholar
  21. 21.
    Motwani, R., Phillips, S., Torng, E.: Non-clairvoyant scheduling. Theor. Comput. Sci. 130(1), 17–47 (1994)CrossRefzbMATHGoogle Scholar
  22. 22.
    Nguyen, K.T.: Lagrangian duality in online scheduling with resource augmentation and speed scaling. In: ESA, vol. 8125 of LNCS, pp. 755–766 (2013)Google Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Authors and Affiliations

  • Spyros Angelopoulos
    • 1
  • Giorgio Lucarelli
    • 2
    • 3
    Email author
  • Nguyen Kim Thang 
    • 4
  1. 1.LIP6Sorbonne Université, CNRS, Laboratoire d’informatique de Paris 6ParisFrance
  2. 2.LIGUniversity Grenoble-AlpesGrenobleFrance
  3. 3.LCOMSUniversité de LorraineMetzFrance
  4. 4.IBISCUniv Evry, Université Paris-SaclayEvryFrance

Personalised recommendations