CAJ | 학술논문

固态硬盘具有高速的随机读取速度、低功耗、体积小等特点,被认为将取代磁盘成为新一代的数据存储设备.但是闪存数据库的查询性能的提高却远小于固态硬盘相比于磁盘I/O性能的提高,其原因在于现有的数据库是基于磁盘设计的,不能充分发挥固态硬盘的高速性能.提出一种名为子连接(sub-Join)的连接算法.首先将数据表的连接列和主键投影为新的子表,然后对子表进行接连操作,最后根据子表的连接结果再从原始数据表中回取查询结果.通过和开源数据库Oracle Berkeley DB的比较实验,结果表明子连接算法比原有算法的性能提高了40%～100%,充分说明了它的优越性.
고태경반구유고속적수궤독취속도、저공모、체적소등특점,피인위장취대자반성위신일대적수거존저설비.단시섬존수거고적사순성능적제고각원소우고태경반상비우자반I/O성능적제고,기원인재우현유적수거고시기우자반설계적,불능충분발휘고태경반적고속성능.제출일충명위자련접(sub-Join)적련접산법.수선장수거표적련접렬화주건투영위신적자표,연후대자표진행접련조작,최후근거자표적련접결과재종원시수거표중회취사순결과.통과화개원수거고Oracle Berkeley DB적비교실험,결과표명자련접산법비원유산법적성능제고료40%～100%,충분설명료타적우월성.
Compared with hard drive disk (HDD), solid state disk (SSD) has a lot of advantages, such as high ran-dom read performance, low power consumption and lightweight form. Therefore it is envisioned to be next genera-tion data storage instead of HDD. However, the enhancement of query performance for flash-based database is not the same as the I/O ratio of SSD to HDD. The reason is existing databases which are designed for HDD can not take full advantage of high I/O performance of SSD. A new join algorithm, Sub-Join, is proposed. Sub-Join firstly pro-jects the column of join and primary key as Sub-Table, and then executes join operations on Sub-Tables. Finally re-sults are gotten from original table according to the result of join on Sub-Tables. The compared experiments with Oracle Berkeley DB show Sub-Join outperforms original indexed nested-loop join at the ratio of about 40%~100%, which shows the high efficiency of this method.