µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
ºÐ»ê Àθ޸𸮠ȯ°æ¿¡¼ ½Ç½Ã°£ 󸮸¦ À§ÇÑ µ¿Àû ½ºÄÉÁ층 |
¿µ¹®Á¦¸ñ(English Title) |
Dynamic Scheduling for Supporting Real-time Processing in Distributed In-memory Environments |
ÀúÀÚ(Author) |
º¹°æ¼ö
ÀüÇö¿í
±è¹Î¼ö
ÃÖµµÁø
À¯Àç¼ö
Kyoungsoo Bok
Hyeonwook Jeon
Minsoo Kim
Dojin Choi
Jaesoo Yoo
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 33 NO. 01 PP. 0076 ~ 0091 (2017. 04) |
Çѱ۳»¿ë (Korean Abstract) |
½Ç½Ã°£ ½ºÆ®¸² µ¥ÀÌÅ͸¦ ó¸®ÇÏ´Â ºÐ»ê Àθ޸𸮠½Ã½ºÅÛÀº ³ëµåÀÇ ºÎÇϸ¦ °í·ÁÇÏÁö ¾Ê°í ÀÛ¾÷ ºÐ¹èÇϱ⠶§¹®¿¡ ƯÁ¤ ³ëµå¿¡ ÀÛ¾÷ÀÌ ÁýÁߵǾî ÀÛ¾÷ÀÇ Áö¿¬ÀÌ ¹ß»ýÇÑ´Ù. º» ³í¹®¿¡¼´Â ½ºÅè¿¡¼ ½Ç½Ã°£ 󸮸¦ Áö¿øÇϱâ À§ÇØ ³ëµå ºÎÇϸ¦ ºÐ»ê½ÃÄÑ ÀÛ¾÷ Áö¿¬À» ¹æÁöÇÏ´Â µ¿Àû ½ºÄÉÁÙ¸µ ±â¹ýÀ» Á¦¾ÈÇÑ´Ù. À̸¦ À§ÇØ ¸Þ¸ð¸®, CPU, ³×Æ®¿öÅ© ºñ¿ë¿¡ µû¶ó ³ëµåÀÇ ºÎÇϸ¦ ÆǺ°ÇÏ°í ºÎÇÏ ³ëµå¿¡¼ ÀçºÐ¹è ´ë»óÀÌ µÇ´Â ÀÛ¾÷À» ¼±ÅÃÇÑ´Ù. ½Ç½Ã°£ ó¸® ÀÛ¾÷ÀÇ Áö¿¬ ¹× ¼Õ½ÇÀ» ¹æÁöÇϱâ À§ÇØ ÀÛ¾÷ÀÇ Ã³¸® ºñ¿ë°ú µ¥µå¶óÀÎÀ» °í·ÁÇÏ¿© ÀÛ¾÷À» ¼öÇàÇÒ ³ëµå¸¦ °áÁ¤ÇÑ´Ù. Á¦¾ÈÇÏ´Â ±â¹ýÀÇ ¿ì¼ö¼ºÀ» ÀÔÁõÇϱâ À§ÇØ ±âÁ¸ ±â¹ý°ú ¼º´É Æò°¡¸¦ ¼öÇàÇÑ´Ù.
|
¿µ¹®³»¿ë (English Abstract) |
Since the distributed in-memory system that processes real-time stream data distribute jobs without considering loads on nodes, jobs are delayed because the jobs are concentrated on one node. In this paper, we propose a dynamic scheduling scheme that prevents work delay by distributing node loads for real-time data processing in storm. The proposed scheme identifies the load of a node according to its status information such as memory, CPU, and network, and selects jobs to be redistributed from the node. We determine the job processing nodes through considering their deadlines and processing costs to prevent the delay and loss of real-time processing. To demonstrate the superiority of the proposed scheme, we compare it with the existing scheme through various performance evaluations.
|
Å°¿öµå(Keyword) |
½Ç½Ã°£ ó¸®
³ëµå ºÎÇÏ
½ºÅè
µ¿Àû ½ºÄÉÁ층
Real-time processing
node load
storm
dynamic scheduling
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|