LNCS 0558 -Job Scheduling Strategies for Parallel Processing

More documents

Recommendations

Info

Series Editors Gerhard Goos, Karlsruhe University, Germany Juris Hartmanis, Cornell University, NY, USA Jan van Leeuwen, Utrecht University, The Netherlands Volume Editors Dror G. Feitelson The Hebrew University Schoolof Computer Science and Engineering 91904 Jerusalem, Israel E-mail: feit@cs.huji.ac.il Larry Rudolph Massachusetts Institute of Technology Laboratory for Computer Science Cambridge, MA 02139, USA E-mail: rudolph@lcs.mit.edu Cataloging-in-Publication Data applied for Die Deutsche Bibliothek - CIP-Einheitsaufnahme Job scheduling strategies for parallel processing : 7th international workshop ; revised papers / JSSPP 2001, Cambridge, MA, USA, June 16, 2001. Dror G. Feitelson ; Larry Rudolph (ed.). - Berlin ; Heidelberg ; New York ; Barcelona ; Hong Kong ; London ; Milan ; Paris ; Tokyo : Springer, 2001 (Lecture notes in computer science ; Vol. 2221) ISBN 3-540-42817-8 CR Subject Classification (1998): D.4, D.1.3, F.2.2, C.1.2, B.2.1, B.6, F.1.2 ISSN 0302-9743 ISBN 3-540-42817-8 Springer-Verlag Berlin Heidelberg New York This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on microfilms or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer-Verlag. Violations are liable for prosecution under the German Copyright Law. Springer-Verlag Berlin Heidelberg New York a member of BertelsmannSpringer Science+Business Media GmbH http://www.springer.de © Springer-Verlag Berlin Heidelberg 2001 Printed in Germany Typesetting: Camera-ready by author, data conversion by Steingräber Satztechnik GmbH Printed on acid-free paper SPIN: 10840923 06/3142 543210
Preface This volume contains the papers presented at the seventh workshop on Job Scheduling Strategies for Parallel Processing, which was held in conjunction with the SIGMETRICS 2001 Conference in Cambridge, MA, on June 16, 2001. The papers have been through a complete refereeing process, with the full version being read and evaluated by five to seven members of the program committee. We would like to take this opportunity to thank the program committee, Andrea Arpaci-Dusseau, Steve Chapin, Allen Downey, Wolfgang Gentzsch, Allan Gottlieb, Atsushi Hori, Richard Lagerstrom, Virginia Lo, Cathy McCann, Bill Nitzberg, Uwe Schwiegelshohn, Mark Squillante, and John Towns, for an excellent job. Thanks are also due to the authors for their submissions, presentations, and final revisions for this volume. Finally, we would like to thank the MIT Laboratory for Computer Science and the School of Computer Science and Engineering at the Hebrew University for the use of their facilities in the preparation of these proceedings. This was the seventh annual workshop in this series, which reflects the continued interest in this field. The previous six were held in conjunction with IPPS’95 through IPDPS’00. Their proceedings are available from Springer-Verlag as volumes 949, 1162, 1291, 1459, 1659, and 1911 of the Lecture Notes in Computer Science series. The last three are also available on-line from Springer LINK. This year saw an emphasis on workloads and their effect, with two invited papers related to heavy tails, and a contributed paper presenting a comprehensive analysis of the workload on a large-scale SMP cluster. We also kept the tradition of having papers that describe real systems, with a much anticipated paper describing the MAUI scheduler. In addition, we saw the maturation of the field as reflected by papers that delved into deeper topics, such as the interaction of scheduling with memory management and communication, the use of SMP clusters, and considerations regarding what metrics to use. We hope you find these papers interesting and useful. August 2001 Dror Feitelson Larry Rudolph
Page 1 and 2: Lecture Notes in Computer Science 2
Page 3: Dror G. Feitelson Larry Rudolph (Ed
Page 7 and 8: Agrawal, M., 11 Bansal, N., 11 Bren
Page 9 and 10: 2 M.E. Crovella 2 Evidence The evid
Page 11 and 12: 4 M.E. Crovella when X is a heavy t
Page 13 and 14: 6 M.E. Crovella delay [43] although
Page 15 and 16: 8 M.E. Crovella 6. Andrei Broder, R
Page 17 and 18: 10 M.E. Crovella 43. R. W. Weber. O
Page 19 and 20: 12 M. Harchol-Balter et al. indepen
Page 21 and 22: 14 M. Harchol-Balter et al. - Given
Page 23 and 24: 16 M. Harchol-Balter et al. Since t
Page 25 and 26: 18 M. Harchol-Balter et al. - There
Page 27 and 28: 20 M. Harchol-Balter et al. 31. S.
Page 29 and 30: 22 A.B. Yoo and M.A. Jette componen
Page 31 and 32: 24 A.B. Yoo and M.A. Jette The DCS
Page 33 and 34: 26 A.B. Yoo and M.A. Jette report o
Page 35 and 36: 28 A.B. Yoo and M.A. Jette A few ne
Page 37 and 38: 30 A.B. Yoo and M.A. Jette The task
Page 39 and 40: 32 A.B. Yoo and M.A. Jette if (the
Page 41 and 42: 34 A.B. Yoo and M.A. Jette Mean Job
Page 43 and 44: 36 A.B. Yoo and M.A. Jette executio
Page 45 and 46: 38 A.B. Yoo and M.A. Jette one of t
Page 47 and 48: 40 A.B. Yoo and M.A. Jette 24. S. P
Page 49 and 50: 42 F. Giné etal. be made taking im
Page 51 and 52: 44 F. Giné etal. Execution time(s)
Page 53 and 54: 46 F. Giné etal. nization grows mo
Page 55 and 56:
48 F. Giné etal. Some studies [24,
Page 57 and 58:
50 F. Giné etal. every time a page
Page 59 and 60:
52 F. Giné etal. Local RQ[∞]:=ta
Page 61 and 62:
54 F. Giné etal. tasks, small and
Page 63 and 64:
56 F. Giné etal. - Slowdown: It is
Page 65 and 66:
58 F. Giné etal. Fig. 6 shows the
Page 67 and 68:
60 F. Giné etal. 5/64 5/32 5/16 5/
Page 69 and 70:
62 F. Giné etal. 30000 25000 20000
Page 71 and 72:
64 F. Giné etal. In this paper, a
Page 73 and 74:
The Influence of Communication on t
Page 75 and 76:
68 A.I.D. Bucur and D.H.J. Epema th
Page 77 and 78:
70 A.I.D. Bucur and D.H.J. Epema by
Page 79 and 80:
72 A.I.D. Bucur and D.H.J. Epema ut
Page 81 and 82:
74 A.I.D. Bucur and D.H.J. Epema In
Page 83 and 84:
76 A.I.D. Bucur and D.H.J. Epema Ta
Page 85 and 86:
78 A.I.D. Bucur and D.H.J. Epema ba
Page 87 and 88:
80 A.I.D. Bucur and D.H.J. Epema Av
Page 89 and 90:
82 A.I.D. Bucur and D.H.J. Epema ex
Page 91 and 92:
84 A.I.D. Bucur and D.H.J. Epema fr
Page 93 and 94:
86 A.I.D. Bucur and D.H.J. Epema 2.
Page 95 and 96:
88 D. Jackson, Q. Snell, and M. Cle
Page 97 and 98:
Page 99 and 100:
Page 101 and 102:
Page 103 and 104:
Page 105 and 106:
Page 107 and 108:
100 D. Jackson, Q. Snell, and M. Cl
Page 109 and 110:
102 D. Jackson, Q. Snell, and M. Cl
Page 111 and 112:
104 B.B. Zhou and R.P. Brent divide
Page 113 and 114:
106 B.B. Zhou and R.P. Brent though
Page 115 and 116:
108 B.B. Zhou and R.P. Brent s l o
Page 117 and 118:
110 B.B. Zhou and R.P. Brent It is
Page 119 and 120:
112 B.B. Zhou and R.P. Brent s l o
Page 121 and 122:
114 B.B. Zhou and R.P. Brent simult
Page 123 and 124:
Effects of Memory Performance on Pa
Page 125 and 126:
118 G.E. Suh, L. Rudolph, and S. De
Page 127 and 128:
Page 129 and 130:
Page 131 and 132:
Page 133 and 134:
Page 135 and 136:
Page 137 and 138:
Page 139 and 140:
Page 141 and 142:
134 Y. Zhang et al. We analyze thre
Page 143 and 144:
136 Y. Zhang et al. Our modeling pr
Page 145 and 146:
138 Y. Zhang et al. Total CPU time
Page 147 and 148:
140 Y. Zhang et al. 3 The Impact of
Page 149 and 150:
142 Y. Zhang et al. Average job wai
Page 151 and 152:
144 Y. Zhang et al. 4. FillMatrix:
Page 153 and 154:
146 Y. Zhang et al. not violate any
Page 155 and 156:
148 Y. Zhang et al. Average job slo
Page 157 and 158:
150 Y. Zhang et al. Average job slo
Page 159 and 160:
152 Y. Zhang et al. The migration f
Page 161 and 162:
154 Y. Zhang et al. improvement is
Page 163 and 164:
156 Y. Zhang et al. Loss of capacit
Page 165 and 166:
158 Y. Zhang et al. 10. B. Gorda an
Page 167 and 168:
160 S.-H. Chiang and M.K. Vernon sc
Page 169 and 170:
162 S.-H. Chiang and M.K. Vernon Ta
Page 171 and 172:
Table 4. Total Monthly Processor an
Page 173 and 174:
166 S.-H. Chiang and M.K. Vernon be
Page 175 and 176:
168 S.-H. Chiang and M.K. Vernon to
Page 177 and 178:
170 S.-H. Chiang and M.K. Vernon Pr
Page 179 and 180:
172 S.-H. Chiang and M.K. Vernon cu
Page 181 and 182:
174 S.-H. Chiang and M.K. Vernon me
Page 183 and 184:
Page 185 and 186:
178 S.-H. Chiang and M.K. Vernon ac
Page 187 and 188:
Page 189 and 190:
182 S.-H. Chiang and M.K. Vernon pe
Page 191 and 192:
184 S.-H. Chiang and M.K. Vernon pr
Page 193 and 194:
186 S.-H. Chiang and M.K. Vernon ch
Page 195 and 196:
Metrics for Parallel Job Scheduling
Page 197 and 198:
190 D.G. Feitelson cummulative prob
Page 199 and 200:
192 D.G. Feitelson Downey has propo
Page 201 and 202:
194D.G. Feitelson LANL-CM5: jobs pe
Page 203 and 204:
196 D.G. Feitelson average response
Page 205 and 206:
198 D.G. Feitelson with very long j
Page 207 and 208:
200 D.G. Feitelson avg bnd-sld 600
Page 209 and 210:
202 D.G. Feitelson 95% confidence i
Page 211 and 212:
204D.G. Feitelson faster convergenc
Page 213:
Agrawal, M., 11 Bansal, N., 11 Bren
show all

LNCS 0558 -Job Scheduling Strategies for Parallel Processing

Create successful ePaper yourself

Delete template?

Save as template?