ࡱ > k p e g R bjbjVV r< r< S= S ` ` ` ` ` D $ 1_ l $% $% $% $% + b 0 1 h ] ] ] ] ] ] ] $ a Od d ] ` "2 * + "2 "2 ] ` ` $% $% ^ P P P "2 ` $% ` $% ] P "2 ] P P r nZ h ^[ $% ,8 0L 4 Z ] _ 0 1_ Z z d dO Z d ^[ ^[ " d ` [ 8 "2 "2 P "2 "2 "2 "2 "2 ] ] P "2 "2 "2 1_ "2 "2 "2 "2 d "2 "2 "2 "2 "2 "2 "2 "2 "2 , & : N{-NbTMNOvZb:gYDn>en{l sepg1 pSpg1 %N NS1 TN8l2 ( 1 . V]'Yf[ 5uP[Oo`f[b, V] b 6 1 0 0 6 5 2 . ѐ]:ghxvz@b, lWS ѐ] 4 5 0 0 5 2 ) Xd : :NMNON{|~Nuv[s|~Y{|WDnvTt)R(ucQZb:gYDnOS>en!jWv^~QZb:gYvhDn:gY~OS{lR M R O 0R M R O {l:gubY~Zb:g>en^Rv^[k~^RۏLOSN-N Q gOv^R\O:N g~vZb:g^R0WNR M R O ۏ NekcQN3 yZb:g>en^RvQOSV{euǏ[[k, bM M B A V{eu\O:N gsOV{eu0Nw~ghfR M R O vk O~vM B F D TM B F H {lf>fMNOpenc-N_vTeO|~YyDn)R(ufTt0 sQ.͋: N{Zb:g>enMNOYDn -NVR{|S: 0 0 0 e.sh_x: A V i r t u a l m a c h i n e m u l t i - r e s o u r c e p l a c e m e n t a l g o r i t h m t o r e d u c e t h e e n e r g y c o n s u m p t i o n i n t h e c l o u d c o m p u t i n g W A N G X i n - j i e 1 L E I Y i n - j i e 1 , Y A N H u a 1 Q I A O Y o n g - q i n 2 ( 1 . C o l l e g e o f E l e c t r o n i c s a n d I n f o r m a t i o n E n g i n e e r i n g , S i c h u a n U n i v e r s i t y , C h e n g d u 6 1 0 0 6 5 , C h i n a . 2 . I n s t . o f Z h e n g z h o u m a c h i n e r y r e s e a r c h , Z h e n g z h o u 4 5 0 0 5 2 , C h i n a ) A b s t r a c t : T o r e d u c e t h e e n o r mous energy produced by the cloud computing system and achieve reasonable utilization of a variety of resources, a virtual machine placing model with multi-resource energy consumption optimization is built and a virtual machine placement algorithmmulti-object resources random multiple sets re-optimization algorithm(RMRO) is proposed. In RMRO, the multi-group sequences of the virtual machine is randomly generated, and after each sequence is optimized , the optimal sequence is selected from the optimized multi-group sequences. Based on RMRO, to optimize the virtual machine allocation sequence , three kinds of policy is proposed. Through the experimental comparison , MMBA is selected as the optimal strategy.Compared to the traditional algorithms which include MBFD and MBFH , RMRO can significantly reduce energy consumption, and make a variety of resources more reasonable in the cloud computing system. Key words: cloud computing ; virtual machines placement ; energy consumption ; multi-resource 0 _ 4O@wN{b/gvSU\TnfSpencTz^ckNLhbP C :gTlQSv gRhV:g?byQ>e0RN{s^SۏLYt A D D I N N E . R e f . { 5 0 D D 0 A 4 C - 7 5 5 7 - 4 F 2 8 - B F D E - 3 F 1 8 6 9 E 2 F C 2 6 } [ 1 ] 0N{/f NyǏs6esNv!j_:N(u7bcOW@xe0s^SToN^(uebv gR[ 2 ] 0Sb7Lk0_o0ŖN0I B M T?̑vN{FUck(WhQNLuT0W^@\penc-N_cON{ gR A D D I N N E . R e f . { 0 C C F 0 E 4 4 - 5 E 7 1 - 4 A 9 D - B B 1 9 - 8 5 5 4 A 2 4 6 F 9 7 6 } [ 2 ] 04O@wpenc-N_W@xeĉ!jv Neib'YTpenc-N_vXYpenc-N_v_N(WŏXRVdkeg͑0 (WNpenc-N_[ؚ;N g$N*NSV N0penc-N_vXYTW@xeĉ!jv Neib'YNu'YϑvN0N{-N_DnRMv NTt bfYvDnm A D D I N N E . R e f . { 0 C A 9 C D 5 6 - 3 B 0 F - 4 5 5 B - B 0 4 1 - C 7 8 2 B C 7 3 C C F 8 } [ 2 - 4 ] 0N{-N_DnSbQX[0C P U 0&^[Tlxv0 (WN{-N'YYpevZb:g>enxvz]\O/fWNgyQRNv NyvhDnvOSُ7h_0Rv gO>enS/fgyQRNv gOel\O0RYyQRNYyDnvOS gHevZb:g>enxvz^[YyDnۏLCgaTOS A D D I N N E . R e f . { C 2 D B 7 4 7 6 - 1 6 0 A - 4 9 2 D - 8 D D 0 - 3 6 9 9 5 6 D A B B E 0 } [ 6 ] 0 WNdkxvzT^zN NyYDnOSZb:g>en!jW0!jW\OST|~vYyDnTtRM g:g~Tb_bZb:g>enYvhDnOS!jW:NMNOcOW@x0(Wdk!jWW@x NcQ NyWNYvhDn:gY~QOSvZb:g>en{l R M R O ( R a n d o m M u l t i - g r o u p R e - o p t i m i z e d ) 0 1 0vsQ]\O VQYf[[N{-N_DnRM NTtTZb:g>envۏLN'Yϑxvz0OYe.s A D D I N N E . R e f . { 3 5 C 1 5 D D B - 1 C 8 4 - 4 C 5 D - B 7 9 1 - 2 2 1 F 6 B E A 4 D 9 9 } [ 2 ] cQv[Zb:g>envM B F D ( M o d i f i e d B e s t F i t D e c r e a s i n g ) {ldk{lOncC P U )R(us[Zb:gۏLM^cR6qTbk*NZb:gRM~Nu g\v;N:g A D D I N N E . R e f . { B 3 6 9 6 A B 7 - 0 0 A A - 4 E 0 1 - B C 8 A - 6 E 5 1 B 3 C 3 6 C 4 5 } [ 2 ] 0FO/f, M B F D *gQvQNDnYlxv0&^[TQX[ [OSNuvq_T0e.s[ 6 ] \YyDn( SbC P U 0QX[0Q~&^[Txv) c6RTOS~Twegg^Zb:gOS>en!jWv^ۏ NekcQWNN~6R>enOSvE A P C {l0E A P C {l(W[Yy{|DnۏLc6RvTeSNf>f0WMNO|~-N_v[ 6 ] 0e.s A D D I N N E . R e f . { A D 7 6 C A 3 0 - 8 0 8 2 - 4 8 F 2 - 9 7 5 F - A 1 A 0 6 D 0 1 E B C 4 } [ 7 ] [Zb:g>en;N:geirtDnvY~'` bY~DnvO(u Ns^aTDnvjm9cQN NyY~zzRy!jWWNdkcQNZb:g>en{lE A G L E , 勗{lSs^aY~Dnv)R(uQ\ЏLvirtppevNMNO0e.s A D D I N N E . R e f . { 7 2 9 C E 4 5 1 - 6 1 E 7 - 4 2 9 C - A 2 6 9 - 1 E 1 2 6 F 2 2 0 5 C D } [ 8 ] cQv{lǑ(uH o l t - W i n t e r c['`Km!jW Km{egKmN NhTgv}v^Ǒ(u NyOSv̀S{lOSZb:gRM;N:gT NybOSe_feH o l t - W i n t e r vSpe