• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ³í¹®Áö C : ÄÄÇ»ÆÃÀÇ ½ÇÁ¦

Á¤º¸°úÇÐȸ ³í¹®Áö C : ÄÄÇ»ÆÃÀÇ ½ÇÁ¦

Current Result Document : 8 / 9

ÇѱÛÁ¦¸ñ(Korean Title) ÀÛ¾÷ ÀÌ·ÂÀÇ ¿äÀÎ ºÐ¼®À» ÅëÇÑ °úÇРŬ¶ó¿ìµå ÀÚ¿ø ¼±Åà ¸ðµ¨
¿µ¹®Á¦¸ñ(English Title) A Science Cloud Resource Selection Model using Factor Analysis of Job Trace
ÀúÀÚ(Author) ±è¼­¿µ   ±èÀ±Èñ   Ȳ¼ø¿í   Seoyoung Kim   Yoonhee Kim   Soonwook Hwang  
¿ø¹®¼ö·Ïó(Citation) VOL 18 NO. 08 PP. 0588 ~ 0596 (2012. 08)
Çѱ۳»¿ë
(Korean Abstract)
Ŭ¶ó¿ìµå ÄÄÇ»ÆÃÀÇ ¹ßÀüÀ¸·Î ´Ù¾çÇÑ °úÇÐ ºÐ¾ßÀÇ ¿¬±¸ÀÚµéÀº ¿¬±¸¿¡ ÇʼöÀûÀÎ ½´ÆÛ ÄÄÇ»Å͸¦ ¿Â-µð¸àµå ¼­ºñ½º·Î Á¦°ø¹ÞÀ» ¼ö ÀÖÀ¸¸ç, µ¿ÀûÀÎ ÀÚ¿ø È®ÀåÀÌ °¡´ÉÇØÁü¿¡ µû¶ó ±×µéÀÇ ¿¬±¸ ȯ°æÀ» È®ÀåÇØ ³ª°¡°í ÀÖ´Ù. ÀÌó·³ »çÀ̾𽺠Ŭ¶ó¿ìµå´Â ÃÖ±Ù ¿©·¯ °úÇÐ ºÐ¾ß¿¡¼­ »õ·Î¿î ¿¬±¸ ȯ°æ Æз¯´ÙÀÓÀ¸·Î ÁÖ¸ñ ¹Þ°í ÀÖ´Ù. ÇÏÁö¸¸ »ç¿ëÀÚÀÇ ¿ä±¸¿¡ ¾Ë¸Â°Ô ½Å¼ÓÇÏ°í µ¿ÀûÀÎ °¡»ó ÀÛ¾÷ °ø°£À» ±¸¼ºÇÏ´Â °ÍÀÌ ¾î·Æ±â ¶§¹®¿¡ ÀÀ¿ëÀÇ Æ¯¼ºÀ» °í·ÁÇÏ¿©, ÀûÀýÇÑ ½ÇÇè ȯ°æÀ» ¹Ì¸® ±¸¼ºÇÏ´Â °ÍÀÌ Áß¿äÇÏ°Ô ¿©°ÜÁø´Ù. ´õºÒ¾î ÀûÁ¤ ¼öÁØÀÇ ¼º´ÉÀ» º¸ÀåÇÏ´Â °¡»ó ¸Ó½ÅÀ» Á¦°øÇÏ´Â ½ºÄÉÁÙ¸µ ¸ÞÄ¿´ÏÁòµµ ¿ä±¸µÈ´Ù.
º» ³í¹®¿¡¼­´Â ÀÛ¾÷ÀÌ·ÂÀÇ Åë°èÀû ºÐ¼®À» ÅëÇÑ »çÀ̾𽺠Ŭ¶ó¿ìµå ÇÁ·ÎºñÀú´× ¸ðµ¨À» Á¦¾ÈÇÑ´Ù. ÀÌ ¸ðµ¨Àº ¾îÇø®ÄÉÀ̼ÇÀÇ ½ÇÇà¿¡ ÀÇÇØ ÃàÀûµÈ ÀÛ¾÷ÀÌ·ÂÀ» ºÐ¼®ÇÏ¿© ±× Ư¼ºÀ» ÆľÇÇÏ°í, °á°ú¸¦ ÀÚ¿ø ÇÁ·ÎºñÀú´× °áÁ¤¿¡ »ç¿ëÇÑ´Ù. ÀÛ¾÷ÀÌ·Â ºÐ¼®À» À§Çؼ­´Â Åë°èÀû ºÐ¼® ±â¹ýÀÎ ÁÖ¼ººÐ ºÐ¼®À» Àû¿ëÇÏ¿´À¸¸ç, À̷νá ÀÛ¾÷ ÇÁ·ÎÆÄÀϵé Áß °¡Àå ¿µÇâ·ÂÀÌ Å« ¿äÀΰú ÀÌ·ÂÀ» ¼±º°ÇÑ´Ù. ¼±º°ÇÑ ¿äÀÎÀº ÂüÁ¶ÇÒ ÇÁ·ÎÆÄÀÏÀ» ¼±ÅÃÇÏ´Â µ¥¿¡ È°¿ëÇϸç, ±× Áß °¡Àå ¿ì¼öÇÑ ¼º´ÉÀ» °®´Â ÀÚ¿ø¿¡ °¡»ó¸Ó½ÅÀ» ¹èÄ¡ ÈÄ ÀÛ¾÷À» ½ºÄÉÁÙ¸µ ÇÑ´Ù. ÀÛ¾÷ ¼öÇà ÈÄ¿¡´Â ±× °á°ú¸¦ ÅëÇØ ÇÁ·ÎÆÄÀÏÀÇ ½Å·Ú¼ºÀ» Æò°¡ÇÏ°í µ¿ÀûÀΠȯ°æ¿¡ ÀûÀÀÇÏ´Â ¾Ë°í¸®ÁòÀÇ À¯È¿¼ºÀ» º¸ÀδÙ. ³í¹®ÀÇ ÈĹݿ¡ Á¦½ÃµÈ ¼º´É ºñ±³·Î½á, Á¦¾È ¸ðµ¨ È°¿ë ½Ã¿¡ ´ë±â ½Ã°£ÀÇ °¨¼Ò¿Í ¼º´É Çâ»óÀ» À̲ø¸ç °¡»óÈ­ ´ÜÁ¡À» ±Øº¹ÇÏ°í Ŭ¶ó¿ìµåÀÇ ÀåÁ¡À» ±Ø´ëÈ­ ÇÒ ¼ö ÀÖÀ½À» º¸¿´´Ù. À̷νá Ŭ¶ó¿ìµå ȯ°æ¿¡¼­ÀÇ È¿°úÀûÀÎ ÀÚ¿ø °ü¸®¸¦ µµ¿ì¸ç °ü¸®¸¦ À§ÇÑ ¿À¹öÇìµå ¶ÇÇÑ ¿ÏÈ­½Ãų °ÍÀ¸·Î ¿¹»óµÈ´Ù.
¿µ¹®³»¿ë
(English Abstract)
The advent of cloud computing makes scientists to extend their research environments over supercomputers to on-demand and dynamically scalable resources. Science cloud has become a trend in various scientific domains these days. Depending on user¡¯s demands, however, it is difficult to supply optimal environment for executing job rapidly and dynamically. Therefore, it is very important to predict user¡¯s requirements and to prepare execution environment in advance. In addition, it needs scheduling mechanisms for virtual machines to offer some level of guaranteed performance of a user application.
This paper proposes a cloud resource provisioning model using statistical analysis of job profiles for science. In this model, we use job profiles which are generated from executions of various applications and identify characteristics of an application by applying statistical analysis. We utilize PCA (Principal Component Analysis) to analyze job profiles and to extract factors which contribute much to execution time or performance. The effective factors are used for selecting reference job profile and deciding a resource site where VM is deployed. An application is executed on the chosen site and its performance result is incorporated into the job profiles with the purpose of evaluating profile¡¯s credit. Performance comparisons with other conditions verify that this model can strengthen strong point of cloud computing and make up for its weakness, since it can offer performance improvements as well as reduction of waiting time. As a result, this model can provide efficient management of cloud resource for a service provider and reduce management overheads on Cloud.
Å°¿öµå(Keyword) »çÀ̾𽺠Ŭ¶ó¿ìµå   ÀÚ¿ø ÇÁ·ÎºñÀú´×   ÁÖ¼ººÐ ºÐ¼®   ÀÛ¾÷ ÇÁ·ÎÆÄÀÏ   Science Cloud   Resource Provisioning   Principal Component Analysis   Job Profile  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå