Á¤º¸°úÇÐȸ ³í¹®Áö D : µ¥ÀÌŸº£À̽º
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
Ä®·³-ÁöÇâ µ¥ÀÌÅͺ£À̽º¿¡¼ ½ºÅ¸ ½ºÅ°¸¶ ÁúÀÇ ¼º´É Çâ»óÀ» À§ÇÑ Á¶Àαâ¹ý |
¿µ¹®Á¦¸ñ(English Title) |
A Join Technique to Improve the Performance of Star Schema Queries in Column-Oriented Databases |
ÀúÀÚ(Author) |
¾È¼ö¹Î
±è°æâ
Su Min Ahn
Kyung Chang Kim
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 40 NO. 03 PP. 0209 ~ 0218 (2013. 06) |
Çѱ۳»¿ë (Korean Abstract) |
Ä®·³-ÁöÇâ(ȤÀº ¿-±â¹Ý) µ¥ÀÌÅͺ£À̽º´Â Å×À̺íÀ» Æ©Çà ´ÜÀ§°¡ ¾Æ´Ñ Ä®·³ ´ÜÀ§·Î ÀúÀåÇÏ´Â »õ·Î¿î °³³äÀÇ µ¥ÀÌÅͺ£À̽ºÀÌ´Ù. ´ë¿ë·® µ¥ÀÌÅ͸¦ ´Ù·ç´Âµ¥ ÀÖ¾î ±âÁ¸ÀÇ Æ©ÇÃ-ÁöÇâ(ȤÀº Çà-±â¹Ý) µ¥ÀÌÅÍ º£À̽ºº¸´Ù ÀûÇÕÇÏ´Ù´Â ¿¬±¸ °á°ú°¡ ÃÖ±Ù¿¡ ¹ßÇ¥µÇ¾ú´Ù. ÇöÀç´Â µ¥ÀÌÅÍ¿þ¾îÇϿ콺 ÀÀ¿ë ºÐ¾ß¿¡ Sybase IQ °°Àº »ó¿ë ¿-±â¹Ý DBMS°¡ ³Î¸® »ç¿ëµÇ°í ÀÖ´Ù. ½ºÅ¸ ½ºÅ°¸¶ ÁúÀÇ´Â µ¥ÀÌÅÍ¿þ¾îÇϿ콺 ÁúÀǷμ »ç½Ç Å×À̺í°ú Â÷¿ø Å×ÀÌºí °£ÀÇ Á¶ÀÎÀÌ Æ÷ÇԵȴÙ. ¿-±â¹Ý µ¥ÀÌÅͺ£À̽º ȯ°æ¿¡¼ »ç½Ç Å×À̺í°ú Â÷¿ø Å×À̺íÀº ±âÁ¸ Çà-±â¹Ý ¶§º¸´Ù »ó´ëÀûÀ¸·Î »çÀÌÁî°¡ Å©´Ù. À̶§ Á¶ÀÎÀº ¸¹Àº µ¥ÀÌÅ͸¦ ó¸®ÇؾßÇÏ´Â ºñ¿ëÀÌ Å« ÀÛ¾÷ÀÌ´Ù. º» ³í¹®¿¡¼´Â ¿-±â¹Ý µ¥ÀÌÅͺ£À̽º¿¡¼ ½ºÅ¸ ½ºÅ°¸¶ ÁúÀÇ ¼º´É Çâ»óÀ» À§ÇØ Á¶ÀÎ ¸ÊÇÎ Æ®¸®¸¦ ÀÌ¿ëÇÑ Á¶Àαâ¹ýÀ» Á¦¾ÈÇÑ´Ù. Á¶ÀÎ ¸ÊÇÎ Æ®¸® ±¸Á¶´Â »ç½Ç Å×À̺í°ú Â÷¿ø Å×ÀÌºí °£ÀÇ Á¶ÀÎ ½Ã ÇÊ¿äÇÑ À§Ä¡ Á¤º¸¸¦ ¹Ì¸® ÀúÀåÇصδ ÀÏÁ¾ÀÇ ¸ÊÇÎ Å×À̺íÀ̸ç ÀÌ ±¸Á¶¸¦ »ç¿ëÇÑ Á¶ÀΠ󸮴 I/O Á¢±Ù Ƚ¼ö¸¦ ÁÙ¿© Àüü ÁúÀÇÀÇ ¼º´É Çâ»óÀ» °¡Á®¿Â´Ù. Á¦¾ÈÇÑ Á¶ÀÎ ±â¹ýÀ» ½ºÅ¸ ½ºÅ°¸¶ º¥Ä¡¸¶Å©¸¦ ÀÌ¿ëÇÏ¿© ÃÖ±ÙÀÇ ¿-±â¹Ý µ¥ÀÌÅͺ£À̽º Á¶ÀÎ ±â¹ýÀÎ ÀκñÁöºí(invisible) Á¶ÀÎ ±â¹ý°ú ¼º´É ºñ±³ÇÑ °á°ú Æò±Õ 1.6¹èÀÇ ¼º´É Çâ»óÀ» º¸¿´´Ù. |
¿µ¹®³»¿ë (English Abstract) |
A column-oriented database is a new database concept that stores tables in column order instead of tuple order. It is known to better handle large data than traditional row(tuple)-oriented databases. In recent years, commercial column-oriented DBMSs, such as Sybase IQ, are widely used in data warehouse applications. In this paper we propose a join technique to improve the performance of star schema queries in column-oriented databases. Our technique uses a join mapping tree structure that pre-stores position information needed to join dimension tables to the fact table in a data warehouse. The use of the join mapping tree decreases the number of I/Os needed to join tables stored in column-oriented databases. A performance comparison with the invisible join technique, the best known join technique for column-oriented database, shows 1.6 times performance gain using the star schema benchmark. |
Å°¿öµå(Keyword) |
Ä®·³-ÁöÇâ(¿-±â¹Ý) µ¥ÀÌÅͺ£À̽º
´ë¿ë·® µ¥ÀÌÅÍ
½ºÅ¸ ½ºÅ°¸¶ ÁúÀÇ
Á¶ÀÎ ±â¹ý
Á¶ÀÎ ¸ÊÇÎ Æ®¸®
ÀκñÀúºí Á¶ÀÎ
½ºÅ¸ ½ºÅ°¸¶ º¥Ä¡¸¶Å©
Column-oriented database
large data
star schema query
join technique
join mapping tree
invisible join
star schema benchmark
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|