在醫(yī)院執(zhí)行模塊的數(shù)據(jù)窗口SQL：

SELECT DISTINCT

ZY_BQYZ.ZYH,ZY_BRRY.BRCH,ZY_BRRY.ZYHM,ZY_BRRY.BRXM,ZY_BRRY.BRXZ

FROM ZY_BQYZ, ZY_BRRY

WHERE ( ZY_BQYZ.ZYH = ZY_BRRY.ZYH )

AND ( ZY_BRRY.CYPB = 0 )

AND ( ZY_BQYZ.SRKS = :al_hsql )

AND ( ZY_BQYZ.SYBZ = 0 )

AND ( ZY_BQYZ.LSBZ = 0 OR ZY_BQYZ.LSBZ = 2 )

AND ( ZY_BQYZ.QRSJ ＜ :ad_today OR ( ZY_BQYZ.QRSJ IS NULL ) )

AND ( ZY_BQYZ.JFBZ = 2 OR ZY_BQYZ.JFBZ = 9 )

AND ( ZY_BQYZ.XMLX ＞ 3 )

AND ( ZY_BQYZ.YZPB = 0 )

ORDER BY ZY_BRRY.BRCH

4小結(jié)

從以上這些例子可以看出，SQL優(yōu)化的實(shí)質(zhì)就是在結(jié)果正確的前提下，用優(yōu)化器可以識(shí)別的語(yǔ)句，充份利用索引，減少表掃描的1/O 次數(shù)，盡量避免表搜索的發(fā)生。其實(shí)SQL的性能優(yōu)化是一個(gè)復(fù)雜的過(guò)程，上述這些只是在SQL語(yǔ)句應(yīng)用時(shí)的一種具體實(shí)例和平時(shí)應(yīng)用中的總結(jié)?？傊瑪?shù)據(jù)庫(kù)的性能優(yōu)化是一個(gè)系統(tǒng)工程，涉及的方面很多。數(shù)據(jù)庫(kù)管理員需要綜合運(yùn)用多種方法，認(rèn)真分析數(shù)據(jù)庫(kù)運(yùn)行過(guò)程中出現(xiàn)的各種問(wèn)題，才能保證數(shù)據(jù)庫(kù)高效地運(yùn)行。