클라이ì-¸íS¸ 서버: 가장 새로운 ì»´í"¨í„° ì‚°ì-...ê³¼ 가장 뜨거운 ì „ë¬¸ì  ìœ í-‰ì-´

장 1				ìœ¼ë¡œ
소개: -				ì
�´Ã¬-¸íS¸				¤.
서버				�€
가장				ì»´í"¨í„°
�´				시ìS¤í...Âœ
ì»´í"¨í„°				ì-ˆì˜
산ì-... 및				ê±´ì¶-ìˆ
가장				을
´ ì „문ì 				ë°-으로
ìœ				�:
í-‰ì-´ì˜				œ 초ì
í-œê°œëS"				ì€
이다.				그것
성격,				�œ
ë°œì „,				� ì¤'ì- (TM)
및				있ëS"
anticipateologies의 depist				자원
수				;
Âœ 대로				�´ 초ì
�´Ã¬-¸íS¸				ì€ 각
Â˜				단ë...
일반ì				˜
ì¸ ì				½ 끝ì-
-의가				지금
ì-†ë‹¤.				í'ë¶€í-œ
˜ 일반				개인
´Ã«S"				�´ë‹¤. PC
�´Ã¬-¸íS¸ì (TM) €				가격
�€				성ëS¥
ì-...무를				´ì˜
�기				또ëS"
위í-˜ì-¬				ì´ë ‡ê²Œ
�크ì-				ìš"인
¼ í-¨ê»˜				2개의
ìž'ë (TM)				매 18
í-˜ëS"				달ì-
�˜Ã«S" ë...¼ë¦¬				의í-˜ì-¬
´ë‹¤				¸°
이다.				것을
�´Ã¬-¸íS¸				기
서버ëS"				ë-Œë¬¸ì-,
ì-„주 ìœ				이 ë (TM)
í-‰				í-¥ì-ëS"
이다.				계속
ë"±ê³¼ ê° (TM)				�«S" 힘이
은,				있다.
다만				초ì
임시ì				ì˜ 이
ì¸				변í (TM) "ëS"
일시ì				사ì-...
 ìœ				구조ì-
�도				있ëS"
¤;				변í (TM)
˜ 기본 ì				"를
ì´ê³				¤: 조직
-œ				��
�¬ÂÂ´Ã«Â‹Â¤				게 í-˜ê³
일반ì				있다, ì
ì¸				-ì±... ìž...ì-ˆ
ìS¹ì¸ì´				권위ëS"
있다;				ªÂ³ 있다,
예를				그리ê³
ë"¤ë©´, Gartner				그것
€ 이				²Œ 된
분ì-¼ì-				ê³¼ì -은
있ëS"				지금
주ìš"				�¤
산ì-...				� 의í-´
°Â€Ã¬ÂÂ¸, ì				이ì „ì-
€ê²ƒì„				ì
�‹Â¤				œê³µëœ
"1995ë...„				ê³¼ì -을
�´Ã¬-¸íS¸				ì œê³µí-
�Œì§€				수
œì˜¤				있다.
�„ 위í-œ				이의
ë (TM)				ê²°í-©í-œ
의ì-´."				íš¨ë ¥ì€
처음				ì±...상ì-
�´Ã¬-¸íS¸				사ì-...ê³¼
서버				ê¸°ìˆ ì
�´ì˜				ì¸ ë (TM)
�„은 ìœ				í-¥
산				�˜ 개인
있ëS"				�«Â‹Â¤. PCëS"
처리				개인
� 직ì ' ì				생산ë
ì´ê³				¥ 및
간ì '				� ì
�¥¼ ì				�¤,
œê³µí-˜ëS"				˜ PC의
¨ ì‹				수ì-
청을				의í-´
�‹Â¤. ê²°ì - ì				ê³±í-Âœ 이
				개체ëS"
�€ ì				또í-Âœ,
'근을,				무ì
�´Ã¬-¸íS¸				-부를
서버				창조í-
공구				수
및 기ìˆ				있다.
의				�´Ã¬-¸íS¸
˜ëS" 미ìˆ				서버ëS"
(TM) ì				이 문ì
œê³µí-˜ëS",				œë¥¼
사ì-...				�«S" 것을
í-„ìš"				ë-ëS"다.
ê´'ì-­				(서버
네íS¸				ì- ì
˜ 진í (TM) "				")
사용				�´Ã¬-¸íS¸ì˜
및				사용
�´Ã¬-¸íS¸				ê³µìœ
서버 ì				자원, ì
„문				ì-S은
ê¸°ìˆ ì˜				개인
€ 이				자원;
매ë Â¥				�´Ã¬-¸íS¸
˜ ì				서버
€ìœ„í-˜				구조ëS"
S"				모ë"
í-©ìž'을				�›¨ì-´ì (TM) €
만ë" 다.				-´
�´ 성과				�„
ì¤'시 ì‹				ê±´ì¶-ê³¼
œ 이				관리
¨ í"„ë¡œì				통ì
œ í- (TM) ìSµ				œì˜
곡ì„				ë°'ì-
을				있ëS"
위로				²Œ í-œë‹¤.
이ë (TM)				€
í-˜ê³ 기				„
ë-Œë¬¸ì-,				�ê³ enterprise-wide
성과				�¼ë¡œ
기대,				개인
ê°€ë (TM)				측ì -을
시간				킨다.
�±Â´ÃªÂ³Â¼				이
ê°€ë (TM)				�� 사ì-...
¸°ë„				ê³¼ì -의
í-˜ê³				순서,
오를				�¸Ã«Â£Â¹
수 있				ë‹¨ê²°ë ¥,
남ì-„				생산ë
있ëS"				¥ 및
��œ 대ì'				±ì„
�€ 있다.				�«S" 것을
그런				ë-ëS"다.
�˜ê²Œ				개인
				í"Œëž˜íS¸í
�¬Â„ϑS", ì				(TM) ˆì´
í-©í-Âœ				�´Ã¬-¸íS¸
ê±´ì¶-ìˆ				
의 ì„				�„을
택				위í-œ
그리ê³				주ìš"
실시ëS"				ê²½ì œì (TM)
²ÂŒ 된다.				€ 기ìˆ ì
실ì				ì¸
œë¡œ				ì›ë (TM) ë
�°€				¼ë„,
ê±´ì¶-ìˆ				�„ 2의
이 이				시ìž'ì-
í"„ë¡œì				í (TM) -인된
íS¸ë¥¼				5개의
위í-´ ë"				ê¸°ìˆ ì
ì ë‹¹í-œ				ì¸
각				Â˜ 단지
�´Ã¬-¸íS¸				¸Ã¬ÂÂ´Ã«Â‹Â¤.
서버				서버
í"„ë¡œì				: -
íS¸ì˜				우리ëS"
¬ (TM) €				서버
� í-˜ëS"				을
기본ì				�¼ ê° (TM) 음
ì¸				ì
�˜ í-œê°œ - "-				�:
2ì¤' 또ëS" 3				�›¨ì-´ê°€
층은?"				„ ì
이다.				œê³µí-˜ëS"
ˆ, 모ë"				ì»´í"¨í„°
성과				í"Œëž˜íS¸í
ì¤'시				(TM) ˆì€
�´Ã¬-¸íS¸				서
서버				„ 위í-´
ì‹				다른
청의 17%ëS"				ê³³ì-
층층				¤Ã­-œë‹¤.
3이ê³ ë (TM)				¤Ã«S" 개인
í-¥ì€ Standish Group				í"Œëž˜íS¸í
International, Inc. 의				(TM) ˆì-
시장				궁극ì
조사				ìœ¼ë¡œ
단체ì-				œë‹¤;
ë"°ë¥´ë©´,				˜
�ê³				¤Ã«S" 또í-Âœ
있다.				다른
ê±´ì¶-ìˆ				서버
은				ì-
�›¨ì-´				„ ì
�³¼ 기ìˆ				�‹Â¤.
�˜ 모ë"				서버
ì-'상ì-				은
�				개인
�¤.				í"Œëž˜íS¸í
�«S" ì‹				(TM) ˆìœ¼ë¡œ
청의				„ 주지
,				ì-SëS"
¬Â²Â˜ ìœ				«S"
í˜-ì-				�„ 통í-´
ê²°ì				¤Ã«Â¥Â¼ ì
-í-˜ê¸° ì				œê³µí-
„ì-,				지도
ì˜ 수,				¤. 거의
�크의				온ê°-
ê·¸ë"¤ì˜				ì»´í"¨í„°
지리ì				í"Œëž˜íS¸í
ì¸				(TM) ˆì€
분산,				서버
본질				으로
및 ì‹				ìž'ë (TM) í-
청의 ì				수
„ë°˜ì				있다.
ì¸				��로,
처리				최ê³
í-„ìš"를				�¬-ì„œ PCì-
ìš"구된				서버
통í-©ê³¼				의
ì¡°í (TM)				많은
�°ì˜				다른
수준				� 및
ê³ ë				많은
¤í-œë‹¤.				가ëS¥í-œ
부ì				, 있다.
ë‹¹í-Âœ				�€ í'œì¤€
ê±´ì¶-				ë...¸ë (TM)
설계				량의
또ëS"				íS¹ì -
�˜ëS"				ì¢...류,
실시ëS"				다른
ì'답				í'ˆì§ˆ
시간				�±Â´, 및
수				�¼ 성과
�. ê±´ì¶-ìˆ				ìS¤íŽ (TM)
의 ì„				íS¸ëŸ¼ì˜
택은				다른
또í-œ ì‹				�„œ
청의				좋다.
개발				
� 및				기ì-...은
미래				다른
± 및 ì				다른
-비				í-„ìš"ì (TM) €
�				기
�¤. 현재				위í-˜ì-¬
문í- (TM)				í"Œëž˜íS¸í
은				(TM) ˆì„ ì„
게				ì -í- 수
�´Ã¬-¸íS¸				있다.
서버				ì„
ê±´ì¶-ìˆ				택의
의 이				이
ì-'상을				폭은
모ë"				�«S" 숫자
�€ ì-SëS"다.				3ì-ì„œ
이				개인
서류ëS"				í"Œëž˜íS¸í
�´Ã¬-¸íS¸				(TM) ˆì-
서버				ëS"
ê±´ì¶-ìˆ				많은
의				서버
기본				ì-
� ì				¤Ã¬- ì
-의í-˜ê³ ,				'근이
2ì¤' 3 층				§Â€Ã«ÂÂ„
ê±´ì¶-ìˆ				¤
을				�다.
기ìˆ				�€ 또í-Âœ
í-˜ê³				지금
ê·¸ë"¤ì˜				ì»´í"¨í„°
각각				통ì‹
이ë"				� 많은
및 ì				다른
œí-œì„				서
œÃ«Â‹Â¤.				다른
�©ì˜				ê³³ì-
개발				¤Ã­-˜ëS"
ë...¸ë Â¥,				„ ì„ 택í-
± 및				지도
¬- 있ëS"				¤ 개인
€ 또í-œ				í"Œëž˜íS¸í
ì-´ë-¤				(TM) ˆì-
í"„ë¡œì				단ë...
íS¸ë"				� 초ì
지를				ì˜
위í-œ ì				¼
í-©í-œ				�‹¤.
ê±´ì¶-ìˆ				�´Ã¬-¸íS¸ì (TM) €
의 ì„				서버
택				í"Œëž˜íS¸í
ë"를				(TM) ˆìœ¼ë¡œ
�¸°				�˜
위í-˜ì-¬				�€
œë‹¤.				�˜Ã¬ (TM) €
장 2				ê³µìœ
ì-­ì‚¬ & defintion: -				자원
ì-­ì‚¬ëS"				사이
�¨ì˜				„
대í- (TM)				다. 각
1996ë...„의				개인
5ì›"ì-ì„œ				í"Œëž˜íS¸í
Oracle ì -부				(TM) ˆì€
금융				ì´ë (TM) í-
(OGF)를				수
‹¤. Solaris 또ëS"				있을
ìœ ë‹‰ìS¤				지도
í (TM) ˜ê²½ì-				모르ê³
그리ê³				다
멀리 IBM				우ì-°í-œ
VMì-				손실
�				�˜
�œ				위í-˜ì-
서				ëS" ë...립ì
žÂ				ì¸
�œ				개인
ì´ë (TM)				�´ë‹¤.
된 UW의				서버
ì¤'í-µ				은
회계 ì				ê³µìœ
œë„.				�„
�€ ê·¸ë-ÂŒ				ªÂ³ , ê³
(ê·¸ë				ì¹˜ê³ ,
‡ì§€				게
ì-S으면				처리í-œ
모ë" )				í (TM)
사ì-...				˜ê²½ì„
�¼ ë" 단				ì
í-˜ë‚˜				�¤.
žÂÃ«Â¡Âœ				�서 ì„
ì´ë (TM)				택í-
í-˜ê³ 단				�¤ 3개의
í-˜ë‚˜				많은
�€Ã«Â²				서버
이ìS¤				ë (TM)
í"Œëž˜íS¸í				일í-œ
(TM) ˆ (ë'˜				기ìˆ
다를				이를
위í-œ				위í-´
Oracle)ì-				�´Ã¬-¸íS¸ì (TM) €
í'œì¤€í (TM)				서버
"í-˜ê¸°				í"Œëž˜íS¸í
위í-œ				(TM) ˆ
다.				¤ë¹„의
ì-„주				ë (TM)
큰				일í-œ
최ì‹				가진
식 Solaris				ë©´ PCì (TM) €
체계ëS"				운영
OGF를				체계)
가진				œë‹¤
ë (TM)				„,
일í-Âœ				¼ ê³µìœ
이				자원
다른				사이
Oracleì-				이
의í-˜ì-¬				구별
œ				만ë"¤ì-´ì
¤ CO				¸ì-¼
�˜ ë...¸ë				í-œë‹¤.
¥ìœ¼ë¡œ				�„œëS", ë (TM)
˜Ã¬-ˆë‹¤.				일í-Âœ
í-‰ì -				기계ëS"
�크, 의ì				개인
-서				í"Œëž˜íS¸í
¬ (TM) € í (TM) œë				(TM) ˆ 및
(TM) ì ì¸				서버
‰				ë'˜
¼ 가진				€ë„
ê·¸				¤
�크를				ë©´ ë (TM)
�œ 막ëS"				료 ì- ë
불을				(TM) 료
ê³				œ; 3.2를
¤ëS"				�œÃ¬Â˜Â¤). ì-¸ì
포í-¨í-œ				œë‚˜,
�크				서버
ë³´ì-ˆ				ì-­í- 은
¬Â²Â˜Ã«S"				ê³µìœ
�˜Ã¬-ˆë‹¤.				�˜
체계ëS"				±
ì € ë³´ì-ˆ				그리ê³
¬Â²Â˜Ã«Â¥Â¼				ì (TM) „ì
¸°				„성을
위í-˜ì-¬				�«S"
˜ÃªÂ³				�¼ 가ì
˜Ã¬-ˆë‹¤.				¸ì˜¨ë‹¤.
€ ê·¸ë (TM)				�´Ã¬-¸íS¸
ì-ˆ 변í (TM)				서버
"í-ˆë‹¤.				�¬-´: -
모ë"				우리ëS"
사ì-...				�´Ã¬-¸íS¸
�¼ 이ë (TM)				서버
í-˜ëS"				�¬-´ë¥¼
위í-˜ì-¬				�¼ ê° (TM) 음
OGF가				ì
지금				�:
1996ë...„ í-Âœ				�´Ã¬-¸íS¸
벌 ì				서버 ì-
€ìª½ì-				이ì...˜
¤Ã«Â¥Â¼ ë"				�›¨ì-´ì˜
계획				�˜Ã«S"
ë (TM) ì-ˆ				�„
�«Â‹Â¤.				�ê³ í-¨ê»˜
주목í-				일í-˜ëS"
만í-˜ê²Œ,				²Œ í-˜ëS"
우리ëS"				œ
(4분기				�›¨ì-´.
œ				�€ 훨ì"¬
˜Ã«S")				�´Ã¬-¸íS¸
¤ Soft/SIS (í- (TM) 생				서버
�)ì (TM) €				ê¸°ìˆ ì˜
가진				가장
를				복잡í-œ
위í-´				다.
¤ Soft/HRMS (인ì 				�„Âœ
자원				ì-
관리				œ 우리ëS"
�€				다른
��로 Oracle가				�
˜				˜
사ì-...을				ìœ ì§€í-
위í-œ				수
이				있다.
주ìš"				그것
Â„ ì				자체
„달í-				�´
수				배부ë
�‹¤ 몇				지도
ë...„을				¤, 그리ê³
ë"°ë¥¼				‰ì´
것을				서로
�‹Â¤. 또í-Âœ				¼
우리ëS"				가지ëS"
다른				�‹Â¤
ì‹ ì²­ì´				기ëS¥ì˜
지ì -ë				많은
ë-Œ				.
�€Ã«Â²				˜ 주ìš"
이ìS¤ë¡œ				지ì-­:
Oracle를				•
것이 ì				�킹
ë‹¹í-Âœ,				¤
그것 Oracle				•
�€Ã«Â²				분산 ì-
이ìS¤				이ì...˜
및 또ëS"				¤
ë°'ì-				•
있ëS"				분산
운영				시ìS¤í...œ
�˜ ë (TM)				관리
일í-œ				•ëS" ì-ˆì
버ì				„을
„으로				�«Â‹Â¤
증ëª...ë				•ëS"
�¼ÃªÂ³				목í'Âœ
ëS" 것이				¼
«Â‹Â¤				�«Â‹Â¤
˜,				•
�«Â‹Â¤. 기ìˆ				
은				˜ì´ìS¤
너무				관리
빨리				•
변í (TM)				인쇄
"í-œë‹¤:				관리
최ì‹				•
식 Solaris				° 관리
체계ëS"				•
ë"				거래
이상				관리
€				• ìž'ì-...
¤.				흐름
�크ëS"				관리
"빨간				숫자 4ëS"
ë°©"ì-				�´Ã¬-¸íS¸
있ëS" í-‰ì -				서버 ì-
체계				이ì...˜
및 교ì				�›¨ì-´ë¥¼
-을				위í-œ
통í-˜ì-¬				이
í-‰ì -				�¬-´
ë¥¼ ê³				�˜ 상ì§-
¤ê¸°				�다.
위í-˜ì-¬				€
˜Ã¬-ˆë‹¤.				�´Ã¬-¸íS¸
˜, í-‰ì -				서버
½ 및 í (TM)				기ìˆ
œë (TM) ì				을
ì¸				ì˜
‰				ì-...무ì- ì
ëS" ê²°ì½"				ë‹¹í-œ ì
지				„ì²´
ì-Sì-˜ë‹¤ -				사ì-...
ì²				�
�«Â‹Â¤.				¸ì- ìž'ë
ì -의:				(TM) í-˜ëS"
�´Ã¬-¸íS¸				²Œ
서버				í-˜ê¸°ì-
�˜ 다량				있ëS"
보도				�¬-´ì˜
«ÂÂ„				�„
�ªÂ³ , 인				‹Â¤.
�´Ã¬-¸íS¸				�€ 기ì-...
서버				˜
œ 무ìS¨ ì				¸ì-
-의의				�¼ 기ëS¥ì
 다량				ì¸
�´ 있다.				˜
�´Ã¬-¸íS¸ì (TM) €				¸ì-
서버ëS"				상호
�›¨ì-´ì (TM) €				�, 그리ê³
¤ë¹„				ì-„마
¼ ì-„ë‹ (TM)				포í-¨í-
´ë‹¤.				지도
�´Ã¬-¸íS¸				¤.
서버				숫자 4
˜ 기본ì				�´Ã¬-¸íS¸
ì¸				서버
�ϑS"				�¬-´
˜ëS"				�´Ã¬-¸íS¸
다른				서버ëS"
�›¨ì-´				�œ 만ë"¤ê³
¼				¤Ã­-œë‹¤: -
�				�´Ã¬-¸íS¸
의í-˜ì-¬				서버
구체ì				체계ëS"
ì¸				복잡í-
ìš"구를				것이
만ë"œëS"				지도
�›¨ì-´				€Ã«Â§ÂŒ, 잘
¼				통í-©
´ì-¸íS¸)를				체계
�‹Â¤.				및 잘
숫자 1ëS"				�˜ëS"
�´Ã¬-¸íS¸				
서버				�œ 기ìˆ ì
교í (TM)				ì¸
˜ì„				은
�‹¤.				ì-ê²Œ
�´Ã¬-¸íS¸				´
ê³¼ì -은				�; 근본ì
				ìœ¼ë¡œ
ìš"구를				ì'용
�¤.				«ÂžÂ¨
서버ëS"				� 및
�를				¤ ì
³ ê·¸				�¥Â¼ 위í-Âœ
후ì-				문ì
ìš"구를				œì´ë‹¤.
ëS" 것을				ê·¸ë"¤ì€
œÃ«Â‹Â¤.				�›¨ì-´
ìš"구를				개발
기				공구ì (TM)
�S",				€ ì „ë¬¸ì
서버ëS"				ì¸
지식				ì-...무를
근원				이
€ë²				을
이ìS¤),				ªÂ³
가공				숨기ëS"
¼				것을
�„				ë-ëS"
)				í-„ìš"로
�±Â°Ã«Â‚˜,				í-œë‹¤.
주변				í-„ìš"로
�¼ 통ì				í-Âœ ì-¸ì
˜, 다른				œë‚˜
Â˜ ì¶"가ì				공구
ì¸				및
ìš"구를				¤Ã¬ÂÂ˜
¼ í-				많은
지도				것은
¤. 다량				ë (TM)
ê±´ì¶-ìˆ				일í-˜,
ì-ì„œëS",				˜ 또í-Âœ
�´Ã¬-¸íS¸ëS"				�´Ã¬-¸íS¸
다수				서버
Â˜				í-„ìš"로
ìš"구를				í-œë‹¤
만ë"¤ê³				�� 있다.
서버ëS"				ì¤'ìš"í-œ
다수				일반ì
�´Ã¬-¸íS¸ë¥¼				ì¸ ì
¤Ã­- 수				ì€
있다.				("shrink-wrapped") ì'용
숫자 1 -				�›¨ì-´,
�´Ã¬-¸íS¸				기ì-...이
서버				«ÂžÂ¨ ê±´ì¶-
거래				¼ í-„ìš"로
�´Ã¬-¸íS¸ì (TM) €				í-˜ì§€ ì-SëS"
서버				ë¥¼
사이				위í-´ ì
�€ ëª...ë ¹				‹¤.
또ëS"
통ì Âœ				�´Ã¬-¸íS¸
�¤ëS"				서버 ì-
것을				이ì...˜ ì
것이				œí'ˆì€
�¤. ì-´ë-¤				지금
교í (TM) ˜ë"				ë„"게 ìœ
œÃ«S",				효í-˜ê²Œ
�´Ã¬-¸íS¸ëS"				되ê³
ìš"구를				있다
ê³				ë©´
서버ëS"				œ
��로				�¤ê³¼
‹¤.				¨Ã¬-´ ì‹ ì²­,
서버ëS"				사ì-...
�´Ã¬-¸íS¸ë¥¼				회계
가진				ì‹ ì²­,
대í (TM)				인원
"를				및
개시í-				ì‹ ì²­).
수				또
ì-†ë‹¤.				다른
�´Ã¬-¸íS¸ì (TM) €				ì¤'ìš"í-Âœ
�€				ë (TM)
�›¨ì-´				í-¥ì€
¼ì´ê¸°				다른
ë-Œë¬¸ì-				공구
ì-´ë-¤ ì				(및 ì-¸ì-´)
í-©í-œ				모ë"ˆë¼
¤ë¹„ë"				ì'용
지ì-				�˜
있을				다른
수				�„ 위í-´
있다.				�¤ëS"
�´Ã¬-¸íS¸				것을
ê³¼ì -은,				이다.
예를				주ìš"
ë"¤ë©´,				€:
�크				•
서버				
¤ë¹„ì-				˜ì´ìS¤:
있을				그래í"½
수				
ìžˆê³ ,				˜ ê±´ì¶-
다른				및 ì-´ë-¤
서버				ì‹ ì²­
¤ë¹„ì-				ë...¼ë¦¬ë"
또ëS" PCì-				지를
달리ëS"				위í-œ
서버				ì-¸ì-´
�ìS¤ì-ì„œ				그리ê³
¼				공구ëS"
�‹Â¤. 다른				ê·¸ë"¤
서ëS",				게 관ë
�´Ã¬-¸íS¸ì (TM) €				‹¤;
서버				면 GUI
ê³¼ì -은				공구ì (TM)
ë (TM)				€ Visual Basic.
일í-œ				•
육체ì				�€Ã«Â²
ì¸				이ìS¤: ì‹
¤ë¹„				ì²­
 있을
수				�€Ã«Â²
있다.
실ì				위í-œ
œë¡œ, prototyping				ì-¸ì-´
서,				그리ê³
�Ã¬-ëS" ë (TM)				공구ëS",
일í-œ PC				파일
-´ì-				시ìS¤í...œ
발í'œ				및
�´Ã¬-¸íS¸				목í'Âœ
그리ê³				ìƒì 
�€Ã«Â²				및 ê±´ì¶-
이ìS¤				ê·¸ë"¤
�€ ë'˜ 다				게 관ë
있ëS"				‹¤;
것을				면
ì„ íƒí-				자료
지도				ì¡°ìž'
¤.				ì-¸ì-´ì (TM) €
서버ëS"				-
ì‹ ì²­				�€Ã«Â²
				이ìS¤ 4GLs.
„ 후ì-				•
S" ë" ì „				ˆìS¤
�„ 위í-œ				로직:
대í˜- ì				
œë„ì-				˜ì´ìS¤ì (TM)
나ì¤'ì-				€
ëS") 이ë (TM)				�€Ã«Â²
í- 수				�„œ
있ê³				�˜Ã¬-´
자료				ë...¼ë¦¬ì
구조				ìœ¼ë¡œ
€ ì (TM) „ì				인 ì‹
„í-˜ë‹¤.				ì²­
�´Ã¬-¸íS¸ì (TM) €
�€ ë (TM)				위í-œ
일í-œ				ì-¸ì-´
있을				그리ê³
수				공구;
�도, 이				ë©´ COBOL.
서류ëS"				•
분산 ì-				분산
�, i.e ê·¸ë"¤				처리:
�ªÂ¸Â°				ì-¸ì-´ì (TM) €
위í-˜ì-¬				공구ëS"
�œ ê±´ì¶-ìˆ				분산
ì-				처리
�´Ã¬-¸íS¸ì (TM) €				그리ê³
�€				모ë"
�˜Ã«S"				상기
육체ì				기ëS¥
ì¸				´ ì „ë¬¸í (TM)
있ëS"				"í-ˆë‹¤ (ì (TM)
ê³³ì- 1				€ 다른
차로				ê³¼í- (TM)
ì-¼ë				ê¸°ìˆ ê³¼
¤í-œë‹¤.				조직
í-´ë¦¬ì-				�¼)
ë"°ë¥´ë©´				œ
(ê·¸ ì (TM) ¸				¼ 위í-´;
),				ë©´
�크의				원격
다른				�ì €
마ë""ì-				호출 (RPC)
³ ê³µë (TM)				공구.
목í'œë¥¼				•
�기				시ìS¤í...œ
위í-˜ì-¬				관리:
í˜'ë ¥í-˜ëS"				�›¨ì-´ì˜
�˜Ã«S"				ì „자
´ 분산				�„ 위í-Âœ
ì-				ë°©ë²-
이ì...˜ì-				및
의í-˜ì-¬				공구,
¬ ¸ 있다.				및
지원				�´Ã¬-¸íS¸
€ 또í-Âœ				서버
분산				�˜ 가ë (TM)
˜ ê³ ìœ í-Âœ				및 ì¡°ì -.
을				이
최ì¢...				˜
ì-ê²Œ				�„은
€ ì-SëS"				�¬-´ì˜
¼ í-œë‹¤.				대ì'
�´Ã¬-¸íS¸
서버				�«Â‹Â¤.
ê±´ì¶-ìˆ				장 6
ì- 있ëS"				�´Ã¬-¸íS¸
�´Ã¬-¸íS¸ëS"				서버
그래í"½				ê±´ì¶-ìˆ :
				ì´ì „
˜ì´ìS¤				¬Â„Âœ 기ìˆ
(GUI)를				된 기ìˆ
�¤,				ì-,
˜,				�´Ã¬-¸íS¸
�´Ã¬-¸íS¸				서버
Â˜ 질량				ê±´ì¶-ìˆ
상í'ˆí (TM)				의
"ëS" GUI				3개의
�´Ã¬-¸íS¸ì˜				ì¢...류ëS" -
í (TM) -산				후ì-
ë-Œë¬¸ì-				봐서 -
큰				� 수
�서				있다.
대략 ì				기본ì
(TM) "다.				ì¸
인쇄 (i.e				�´Ã¬-¸íS¸
�크				서버: -
인쇄				기본ì
ê°ê³ ëS"				ì¸
또ëS"				�´Ã¬-¸íS¸
발í'œ (i.e X				서버
ì°½)ì (TM) €				ê±´ì¶-ìˆ ,
ê° (TM) 은				개인
�´Ã¬-¸íS¸				ì‹
서버				� - 2개
시ìS¤í...œ				로
지원				��¤: 개인
ë†'게 íS¹ì				í"Œëž˜íS¸í
-í-œ 기ëS¥				(TM) ˆì-
¤. 이				�´Ã¬-¸íS¸
íS¹ìˆ˜				부속,
ëª©ì 				및
€ ì¤'ìš"í-œ				서버
ë (TM) ì-ˆ,				ì-
이				서버
서류ëS"				부속.
기ëS¥ì-				후자ëS"
있ëS"				�œ ê³µìœ
±ì„				자원
				파일
�œ				¤Ã¬ (TM) € ê° (TM)
�´Ã¬-¸íS¸				은
서버				인쇄
ê±´ì¶-ìˆ				¤,
ì-				�€Ã«Â²
게				이ìS¤,
다				또ëS"
장 3				ì-´ë-¤ application-specific
�´Ã¬-¸íS¸				«Â‹Â¤.
Â˜ 의미: -				기간
�´Ã¬-¸íS¸				�´Ã¬-¸íS¸ì (TM) €
Â˜ 사ì-...				서버ëS"
의미: -				-´
�´Ã¬-¸íS¸				í"Œëž˜íS¸í
서버ëS"				(TM) ˆ 및
사ì-... ì				ì'용
-ë³´				�›¨ì-´
�˜				�¼
				"
¬Â„Âœ				�œÃ«Â‹Â¤
다음				�œ ì-½ê°„ ì-
´ê³				매í-˜ê²Œ).
일반ì				기본ì
ìœ¼ë¡œ				ì¸
				�´Ã¬-¸íS¸
�œÃ«Â‹Â¤.				서버
�€ 혁ì‹				ê±´ì¶-ìˆ
의				은
�„œ 누ì				숫자
ëœ				9ì-ì„œ
„				�다
‚´ëS"				(이미
숫자				숫자
1ì-ì„œ				6ì-ì„œ ë"
�다.				ˆ
ˆìS¤				ë³´ì´ê³ ).
�Â€				숫자 9
일괄				기본ì
¼ 가진 1960				ì¸
�¬Â„Âœ				�´Ã¬-¸íS¸
�‹¤. 1970				서버
ë...„대ì-				ê±´ì¶-ìˆ
있ëS"				기본ì
주ìš"				ì¸
혁ì‹				�´Ã¬-¸íS¸
은				서버
¸				€
ì...˜ 처리				�
(OLTP) 2 ë°"íƒ- í (TM)				�
"ë©´ì- ì				¬ 통ì‹
-ë³´ 기ìˆ				망)의
�¼ 가ì				 일반ì
¸ì˜¨,				ìœ¼ë¡œ
�ªÂ²ÂŒ				œë‹¤. ì
�°Ã­-Âœ				„ì²´
�„ 사ì-...				ì§'í-©ì-
ê³¼ì -의				의í-˜ì-¬
�Â´ÃªÂ³ .				PC-LAN로
ì¡°í-©ì-				보통
있ëS"				기ìˆ
일괄				ë˜ê³ ,
처리				í-œê°œ
그리ê³				이상
OLTPëS"				ê³µë (TM) PC
최대				(서버
기ì-...의				) í"ŒëŸ¬ìS¤
ì -ë³´				개인
�˜				사용
ì¤'í-µì-				(개인
있ëS"				í"Œëž˜íS¸í
것을				(TM) ˆ)를
다. ê·¸				위í-œ
ë-Œ				많은
1980ë...„대				PC가,
개인				¬ ¸ 있다.
측ì				이ì-
�œ				지ì-­
„ 보편ì				서버
ìœ¼ë¡œ				í"Œëž˜íS¸í
기ì-...을				(TM) ˆì€
통í-˜ì-¬				ê·¸ë"¤ì-
´Ã«ÂÂ¼ÃªÂ³ ì				enterprise-wideì (TM) € ì (TM)
�˜ 및				¸ë¶€
ëS" í-œ ì (TM)				�크로,
"다.				그리ê³
지금 1990				 PC LANs 보통
�¬Â„Âœ,				를 ì
�´Ã¬-¸íS¸				�¤.
서버ëS"				�€ 숫자
ì -ë³´				10ì-ì„œ
�˜				�다.
�˜Ã«S"				PCì (TM) € PC
부분				LANs의 ì
후ì-				ì-ì„œ 1
				차로
ªÂ³				í'œí˜„í-´,
일반ì				이
ìœ¼ë¡œ				기본ì
í-¨ê»˜				ì¸
œë‹¤. ì				�´Ã¬-¸íS¸
€ê²ƒì€				서버
˜ ì-­í- 및				개ë...
˜				숫자 10ëS"
�´ë‹¤.				ì „í˜-ì
숫자 1				ì¸ PC-LAN
œ				온ê°-
�´Ã¬-¸íS¸				ì»´í"¨í„°
서버				및
이러í-œ				�크
상í (TM)				ë©´ PC, ìœ
©ì-ì„œ				닉ìS¤,
�´Ã¬-¸íS¸				ªÂ²Â©, LANs 및
Â˜ 사ì-...				WANs)ì- ì
€ (또ëS"				ìš©
�´Ã¬-¸íS¸				‹¤.
서버)				기초 ì
사ì-... 및				€ìª½ì-: -
ê¸°ìˆ ì				기본ì
ì¸				ì¸
통찰ë Â¥				�´Ã¬-¸íS¸
및				서버 ì
매매				€ìª½ì- -
�의 모ë"				�ªÂ³ , í˜'ë Â¥
ë°©ë²-ì-				�ªÂ³ ë...립
ì œí'ˆì˜				가공 ë
거의				(TM) 료 ì-
각				ë (TM)
ì¢...류ì-				료가
ì				있다.
" 대ì¤'ì				기간 ë
ì¸ ìœ				(TM) 료 ì-
ëª...				ë (TM) 료
�, 및				�€ 서버
.				이 ì-†ê³ ,
�€ íS¹ì -í-Âœ				ì‹
�˜				청의
„				서버
배수í-´				´ 개인
ê²½í-¥ì´				í"Œëž˜íS¸í
있다;				(TM) ˆì-
˜ ê·¸ë				있ëS"
‡ê²Œ				„
써, 실ì				"
œë¡œ				�œÃ«Â‹Â¤. 이
˜ 가깝				기초ì-
보편ì				ìž'ë (TM)
ì¸				í-˜ëS"
�을 í (TM)				�크ëS" ë (TM)
.				료 ì- ë
�´Ã¬-¸íS¸				(TM) 료
Â˜ 기ìˆ ì				�크로
ì¸				¤.
의미: -				�€ ìž'은 PC
이í-´				LANs, ë"±ë"±
�´Ã¬-¸íS¸				ê°'이
¥Â¼ 위í-Âœ				싼
ìœ 용í-Âœ				�«Â‹Â¤;
출발ì				˜
ì€ Gartner				�˜Ã«S"
´				서버
ì-½ì‹ ì				의
�¤:				€ 체계
´ì-¸íS¸				ì (TM) „ì
서버				„성을
�˜Ã«S"				œí‚¤ê³
�¬-				시ìS¤í...Âœ
ëS"				관리
�Â´Ã«Â‹Â¤ ì‹				ì-´ë
ì²­, ê·¸ì¤'				«Â¡Âœ
�				이끌ì-´
�¬ÂÂ´Ã«Â‹Â¤				낸다.
í'€ê·¸ë¦´				기간
�í...Œì´ì...˜				í˜'ë ¥
ë©´ PC)."ëS"				�€ ì'용
이 ì				�›¨ì-´ê°€
-의ëS"				œ ì -상
�´Ã¬-¸íS¸				�˜Ã«S"
�€				서버
œ 계산				인
그리ê³				„
�›¨ì-´				"
구조 (ì‹				�œÃ«Â‹Â¤,
청은				상호
�˜Ã«S"
�¬-				�´Ã¬-¸íS¸ì (TM) €
§Â€Ã«ÂÂ„				서버
¤)				끝은
				서버
관í-˜ì-¬				ì- ë'˜ 다
��¤ 다ëS"				있다.
것을				�€ ë (TM)
¤.				일í-œ ì‹
€				청의
�´Ã¬-¸íS¸				부분
�€ 온ê°- ì				사이
-ë³´				�˜Ã«S" ì‹ ì²­
�œ¼ë¡œ				사이
PC를				상호
" ë°©ë²-				�, 다만
이다				�‹Â¤. 기간
생ëª... ì				ë...립
ì„				�€ ì‹
�‹Â¤.				청의
ì				모ë"
„갈의				�´ 1개의
�˜ 발생:				í"Œëž˜íS¸í
호ìS¤íS¸				(TM) ˆ (보통
기본				개인
ê±´ì¶-ìˆ				í"Œëž˜íS¸í
´ì-¸íS¸				(TM) ˆ)ì-
서버				있ëS"
ê±´ì¶-ìˆ				„
ì-„ë‹ (TM)				"
니다):				�œÃ«Â‹Â¤.
ªÂ²Â©				부속
�›¨ì-´				사이
� 모ë" ì				ì-´ë-¤
-ë³´ëS" ì¤'ì-				�´Ã¬-¸íS¸
(TM) 주ì				서버
„산기				„ ì (TM)
ì-ˆì-				¸ë¶€ì-
있다.				€ ì-S다.
ëS"				¤Ã¬ÂÂ€
ë¶ (TM) 잡ê³				또í-œ
ì € ì				�¬Â§Â€
-보를				다른
보내ëS"				의미ì (TM)
�„				€ í-¨ê»˜
				교í (TM) ˜í-
¼ 상호				수 있
‹¤.				기간 ë
ªÂ²Â©				(TM) 료 ì-
�›¨ì-´				ë (TM) 료ì (TM)
구조ëS"				€ í˜'ë ¥
-´				�„,
í"Œëž˜íS¸í				그리ê³
(TM) ˆì-				다.
�€ ì-SëS"다.				�€
 상호				¼
� PCì (TM) € ìœ				오í-´ë¥¼
닉ìS¤				�¤ëS"
�„				´ 된다.
ì-¬ 끝ë‚				또í-œ
수				서버ì
있다.				(TM) € ê° (TM)
ªÂ²Â©				은
�›¨ì-´				�¬Â§Â€
˜ ì				다른
œí-œì€				보다 ì
ê·¸ë"¤ì´				ê²Œ ìœ
쉽게				ëª...í-œ ì
그래í"½				-립이
				또ëS"
¼				 및
�지				� 또ëS"
ì-S거나				 있다.
지리ì				모ë"
ìœ¼ë¡œ				주ìš" ì
				-립은
�서				숫자
다수				11ì-ì„œ
�€Ã«Â²				í-¨ê»˜
이ìS¤ì-				�다.
ì				œ 기본ì
ì-SëS"다				ì¸
이다.				�´Ã¬-¸íS¸
지난				¥Â¼
�ì-ì„œëS",				�«S"지
ªÂ²Â©Ã¬ÂÂ€				ì-´ëS
Âœ				�¼ÃªÂ³ ì-
�´Ã¬-¸íS¸				의í-˜ì-¬)
서버				숫자 11
ê±´ì¶-ìˆ				¬ÂÂ˜
ì- 있ëS"				�´Ã¬-¸íS¸
¡Âœ				서버
�´				시ìS¤í...Âœ
„				˜ 각ì¢... ì
ˆÃ«Â‹Â¤				-립은,
� 통ì‹				많은
망				¤
파일				�ªÂ²ÂŒ
ê³µìœ				�´Ã¬-¸íS¸
ê±´ì¶-ìˆ				Ã¬Â„Âœ
´ì-¸íS¸				다른
서버				� 분í (TM)
ê±´ì¶-ìˆ				"í-œë‹¤ (.
ì-„ë‹ (TM)				�€
니다): -				모ë'ê°€
원래 PC				1개의
�크ëS"				구조
�€ ê³µë (TM)				내의
�서				´Ã«Â‹Â¤
�© í (TM)				생ëª... ì
˜ê²½ì-				ì„
„				ì-´ë'¡ê²Œ
œÃ«"ϒ-˜ëS"				í-œë‹¤:
파일				�´Ã¬-¸íS¸
ê³µìœ				서버
ê±´ì¶-ìˆ				ê±´ì¶-ìˆ .
ì-				€ 또í-Âœ
«Â‹Â¤.				¸Â° 쉬운
ìš"구된				��œë¡œ
 일은				�´Ã¬-¸íS¸
�© í (TM)				서버
�œ ê·¸ ë-Œ				(의미
(를				기본ì
�¬ ë...¼ë¦¬				ì¸
및				�´Ã¬-¸íS¸
자료)				서버)ëS"
�¤. 파일				�«Â‹Â¤
ê³µìœ				이라ëS"
ê±´ì¶-ìˆ				�˜
은				이끌ì-´
ê³µë (TM)				ë‚´ê³ ,
경우ì-				í˜'ë ¥
-이				¼ ê° (TM) 은
�´, ê°±ì‹				다른
ë...¼ìŸ				ê¸°ìˆ ë¡œ
ë‚(R)다				대ì‹
ìž'ë (TM)				되ê³
í-˜ê³ ,				있다.
				일반ì
�‰Ã¬ÂÂ€				ì¸
ë‚(R)다. 1990				�´Ã¬-¸íS¸
�¬Â„ϑS",				서버
파일				ê±´ì¶-ìˆ : -
ê³µìœ				기본ì
의				ì¸
�‰Ã¬ÂÂ´				�´Ã¬-¸íS¸
ë-Œë¬¸ì-				서버ì
¸				(TM) € 3.1ì (TM) €
ì˜ 수				3.2ì- 있ëS"
�기				모ë" ì
˜ÃªÂ¸Â°				-립의
ë-Œë¬¸ì-				기본ì
ë°"뀐 PC				ì¸ ì
� 통ì‹				œí-œì€
망				ê·¸ë"¤ì´
¬				-´
�)				�
�€ (단지				«S" ë°©ë²-ì-
대략				있ëS"
12ëª...의				�›¨ì-´
ë§Œ ë (TM)				„ ì
시로				� 이다.
�œÃ­Â‚¬ 수				�€,
있다)				€
그래í"½				�›¨ì-´
				또ëS"
˜ì´ìS¤ (GUIs)				¤ë¹„ê°€
대ì¤'ì 				의í-˜ì-¬
되ê³				�´Ã¬-¸íS¸ì (TM) €
�ªÂ²Â©ÃªÂ³Â¼				서버
끝 ì				�„´ë‹¤ëS"
„시를				것을
�„Âœ				�œ ì-
¸				매í-˜ë‹¤.
¼				이 ì ϒ-Âœ
­Ã¬Â‹ÂœÃ¬Â˜Â¤).				및 ì-
PCëS" 지금				�„Âœ
�´Ã¬-¸íS¸				�기
서버				�S",
ê±´ì¶-ìˆ				�›¨ì-´ì-
ì-ì„œ				있ëS"
˜ê³				�´Ã¬-¸íS¸
있다				서버
·				관계ëS"
�´Ã¬-¸íS¸				�´Ã¬-¸íS¸
서버				또ëS"
ê±´ì¶-ìˆ : -				¡Âœ
이				�›¨ì-´
위í-œ				위치
목í'œëS"				그리ê³
ê³¼ì -의				근본ì
�€Ã«Â¥Â¼				ì¸
ë°'ì-				¤ë¹„의
있을				ì-´ë-¤
ë°°ê²½				분류ë"
�˜				지은
¼				�Âœ ì
기				�ì-¼ í-œë‹¤.
위í-œ				근본ì
�¤. 기ìˆ				ì¸ ì -í (TM)
의				"ëS"
많은				�´Ã¬-¸íS¸ì (TM) €
�„œëS"				�€
�€				¤ÃªÂ°Â€
˜ 현재				(각각)
�œ í-­ìƒ				˜ê³ ì
�˜Ã«S"				" ì-­í-
것과 ê°				이다
(TM) 이				이다,
„ 기ìˆ				이 ì-­í-
ì-»ëS"다.				은
˜,				자치
기ìˆ				빌ë"©
ì-ëS"				ë¸"록
다른				사이
ì-´ë-¤				서
자ì-° ì				�¤. 그런
ì´ê³				서ëS",
�ìS¤ëŸ¬ìš´				�Ã¬ÂÂ˜
현상				í-œê°œëS"
ê° (TM) 이				¤ (그것
� 다만				있다
있다.				�´Ã¬-¸íS¸ ì-­í-
€				이)를
· 및				다 또
ì›"ë"œ ì (TM)				다른
€ì´ë"œ				í-œê°œëS"
웹을				¤ (서버
위í-´				ì-­í- 이
ì´ë ‡ê²Œ				¼ ì
이다.				œê³µí-˜ê³ .
이				�€
면담 ë				�´Ã¬-¸íS¸
(TM) ì-ˆì-,				서버
우리ëS"				´ë‹¤. 큰
				ìœ ì-°í-œ
				설ì -은
·ì„				이
만ë"
과거 30				개ë...ì˜
ë...„				ì¡°í-©ì-
내내				의í-´
� ì-´ë-¤ì˜				ìŒ"ì-„
ì¤'ìš"í-œ				올릴
„				수
� ë³¼				있다.
�¤. 이				�€ 숫자
�를				12ì-ì„œ
ê²€í† í-œ				�다.
후ì-,				�
우리ëS"				½Ã¬-ì„œ
·ì„				‚˜ëS"
�ëS"				것과 ê°
근본ì				(TM) 이
ì¸				�´Ã¬-¸íS¸
ê¸°ìˆ ì˜				서버
2개를				ê±´ì¶-ìˆ
볼				의
�¤.				숫자 12
¸				원리ëS",
˜				빌ë"©
몇백ëª...				ë¸"록
ì-¼ë"Âœ				¤Ã¬ÂÂ˜
ì-ˆì-				
있ëS"				그리ê³
�«Â¥Â¼				� ë'˜
¸Â°				€ë„
위í-˜ì-¬				¤.
�œ 가장				��로,
널리				�«S"
퍼진				�´Ã¬-¸íS¸ì (TM) €
통ì‹				서버
망의				ì-­í- 이
�íS¸,				있ê³
원래				Â„ 다른
�				빌ë"©
�				가진
¬ 통ì‹				많은
망) 기ìˆ				�´Ã¬-¸íS¸
및				서버
�� í-œê°œ				참가í-
이다. ì				지도
Âœ2ëS" TCP/IP 의				¤. ê³ ë ¤ëœ
�를				íS¹ì -í-œ
�게				˜
�ê³				�Œ
�ªÂ²ÂŒ				�´Ã¬-¸íS¸
�ªÂ¸Â°				또ëS"
위í-˜ì-¬				서버
ì				이다.
„세계				�›¨ì-´ì-
있ëS"				있ëS"
�«Â¥Â¼				�´Ã¬-¸íS¸
²Œ í-˜ëS"				서버
�›¨ì-´				ê±´ì¶-ìˆ
�´Ã«Â‹Â¤.				의
·ì„ í†				현실í (TM)
ë¡ í-œ				"ëS" 숫자
후ì-,				12ì-ì„œ
우리ëS"				€ ì-SëS" ()
ê·¸ ë-ÂŒ				«ÂžÂ¨ ì-¸ì-´
ì›"ë"œ ì (TM)				및
€ì´ë"Âœ				�¬-´ë¥¼
웹ì-				통í-´
˜ 주의,				이다.
�¼ 돌				�˜Ã«S"
�¤.				�¬ÂÂ˜
�€ 1대의				�크가
�¬-				�´Ã¬-¸íS¸
달리ëS"				서버
�´Ã¬-¸íS¸				ê±´ì¶-ìˆ
«ÂžÂ¨Ã¬ÂÂ´				의
ì-½ê°„				육체ì
íS¹ì -í-œ ì				ì¸
-보를				현실í (TM)
¸Â°				"ì-
위í-˜ì-¬				의í-˜ì-¬
또				¬ ¸ 있다;
다른				í-„ì-°ì
í-œê°œì-				ìœ¼ë¡œ
달리ëS"				기간
서버				�´Ã¬-¸íS¸
«ÂžÂ¨ÃªÂ³Â¼				서버ëS"
S"				분산
˜ 기본ì				¼ 위í-Âœ
ì¸				ë (TM)
�´Ã¬-¸íS¸				되ì-´
서버				ê²½í-¥ì´
ê±´ì¶-ìˆ				있다.
의				�´Ã¬-¸íS¸
ê²€í† ë¡œ				서버
또ëS"				ê±´ì¶-ìˆ
ì-½ê°„				은
¤Ã«S"				기ìˆ 의
다				다른
�다.				ì-´ë-¤ íS¹ì -
웹은 ì				PC, 또ëS"
-보의				�
�€Ã«Â¥Â¼				„œë§Œ
¸Â°				부수ì
위í-˜ì-¬				ìœ¼ë¡œ
웹				이다.
�°Ã¬ €				˜, 현재
´ì-¸íS¸)가				상í (TM)
각ì¢...				©ì-ì„œ,
웹				�´Ã¬-¸íS¸
서버ì				�€ PC 기ìˆ
(TM) €				�˜ ì
S"				ì-ì„œ
�´Ã¬-¸íS¸				주로 ì
서버				�‹¤ 보통
ê±´ì¶-ìˆ				ì
을				�
ì-¬ 또ëS"				�� 의
«ÂžÂ¨Ã¬ÂÂ€				위ì-
Â˜ 일반				1.2ì-ì„œ
				시ìž'í-œ)
Ã¬-­ (CGI)를				Gartner ì
				œ 것과
다				ê° (TM) 이.
˜ì-ˆë‹¤. Internet				자치
WWW가				�´Ã¬-¸íS¸
일반ì				서버
ì¸				ê±´ì¶-ìˆ
계산				의 이
그리ê³				일반ì
통ì‹				ì¸
�가 되ëS"				모ì-'은
ë-Œ, 이				(의
준ì-„í-Âœ				�´Ã¬-¸íS¸
�´Ã¬-¸íS¸				서버
서버				관계,
관계ëS"				�´Ã¬-¸íS¸ ì-­í- ,
í (TM)				서버
있다.				ì-­í-
1개의				건물
그런 í (TM)				막ëS"다)
-장은				개방í˜-
				í"„ë
‹Â¤. ê³¼ì				­ÂÂ¬ ì‹ ì²­
-ì-ì„œ				ê±´ì¶-ìˆ
나ì¤'ì-				의
이 ì‹				기본ì
관í-˜ì-¬				ì¸
ë"				Â´Ã«Â‹Â¤.
많은				Assumptions:-
것,				1.One client is connected to at most one server at a
˜				time. [The customer later refuted this assumption.]
�„,				2.Replication is a secondary effect of the existing
우리ëS"				fat-client architecture; we assume that updates to
웹의				one server are automatically propagated in a timely
ê³ ì „ì				fashion.
ì¸				3.A single client may have more than one session.
�´Ã¬-¸íS¸				[Replaced Assumption 1.]
서버				4.All calculated columns (columns that represent
�- ì§'ì¤'í-				behavior rather than aspects) are easily & quickly
�¤. ì-¸ì-´ 웹				calculated on the server.
�´Ã¬-¸íS¸ì (TM) €				5.Deletion or insertion of a row forces a window
서버ëS"				update on the client.
서로				6.Transmission of client-server traffic is out of
¤ HTTP				scope.chapter-7
í...ìS¤íS¸ ì				Architecture Types:- When considering a move to
„송				client/server computing, whether it is to replace
í"„ë¡œí†				existing systems or introduce entirely new systems,
�				practitioners must determine which type of
. 당ì‹				architecture they intend to use. The vast majority of
은 HTTP를				end user applications consist of three components:
�ˆ 배울				presentation, processing, and data. The client/server
�„				architectures can be defined by how these
�¤,				components are split up among software entities and
˜ 당ì‹				distributed on a network. There are a variety of
은				ways for dividing these resources and implementing
기본ì				client/server architectures. This paper will focus on
ì¸ HTTP				the most popular forms of implementation of
«ÂžÂ¨Ã­-˜ëS"				two-tier and three-tier client/server computing
CGI를				systems. Two-tier Architecture:- Although there are
í-˜ê¸°				several ways to architect a two-tier client/server
위í-˜ì-¬				system, we will focus on examining what is
�를				overwhelmingly the most common implementation. In
�¬-¼ í-˜ê³				this implementation, the three components of an
˜ 근본ì				application (presentation, processing, and data) are
ì¸ ì² í- (TM)				divided between two software entities (tiers): client
및				application code and database server (Figure 2). A
˜				robust client application development language and a
기본í˜-				versatile mechanism for transmitting client requests
및 기ëS¥				to the server are essential for a two-tier
¼ í-œë‹¤.				implementation. Presentation is handled exclusively by
장 4				the client, processing is split between client and
ê³¼ì -: -				server, and data is stored on and accessed via the
�´Ã¬-¸íS¸				server. The PC client assumes the bulk of
ê³¼ì -: -				responsibility for application (functionality) logic with
�´Ã¬-¸íS¸				respect to the processing component, while the
이다				database engine - with its attendant integrity checks,
ê³¼ì -				query capabilities and central repository functions -
°Â€ ì-...무				handles data intensive tasks. In a data access
�¤)를				topology, a data engine would process requests sent
„´ë‹¤ëS"				from the clients. Currently, the language used in these
것을				requests is most typically a form of SQL. Sending
				SQL from client to server requires a tight linkage
«ÂžÂ¨)				between the two layers. To send the SQL the client
서버				must know the syntax of the server or have this
�ìS¤				translated via an API (Application Program Interface).
�«ÂžÂ¨)ì-				It must also know the location of the server, how
�를				the data is organized, and how the data is named.
보내ëS".				The request may take advantage of logic stored and
�´Ã¬-¸íS¸				processed on the server, which would centralize
«ÂžÂ¨Ã¬ÂÂ€				global tasks such as validation, data integrity, and
보통 ì‹				security. Data returned to the client can be
청의				manipulated at the client level for further sub
				selection, business modeling, "what if" analysis,
Ã¬-­				reporting, etc.
�„				Figure 2 - Data Access Topology for two-tier
ªÂ³ ,				architecture. Majority of functional logic exists at the
ê°€ ìž...ë				client level The most compelling advantage of a
¥í-œ				two-tier environment is application development
¼ ìœ				speed. In most cases a two-tier system can be
효í-˜ê²Œ				developed in a small fraction of the time it would
í-˜ê³ ,				take to code a comparable but less flexible legacy
서버				system. Using any one of a growing number of
«ÂžÂ¨Ã¬-				PC-based tools, a single developer can model data
ìš"구를				and populate a database on a remote server, paint a
�ªÂ³ ,				user interface, create a client with application logic,
ë-Œë-Œë¡œ				and include data access routines. Most two-tier tools
ˆìS¤				are also extremely robust. These environments
�„				support a variety of data structures, including a
‹¤.				number of built in procedures and functions, and
�´Ã¬-¸íS¸				insulate developers from many of the more mundane
근거í-œ				aspects of programming such as memory
ê³¼ì -은				management. Finally these tools also lend themselves
ê°€				well to iterative prototyping and rapid application
ë³´ê³				development (RAD) techniques, which can be used
상호				to ensure that the requirements of the users are
S" ì‹				accurately and completely met. Tools for developing
청의				two-tier client/server systems have allowed many IS
ì„				organizations to attack their applications backlog,
�´Ã«Â‹Â¤.				satisfying pent-up user demand by rapidly developing
�´Ã¬-¸íS¸				and deploying what are primarily smaller
ê³¼ì -은				workgroup-based solutions. Two-tier architectures
í-´ê²°ì±...				work well in relatively homogeneous environments
íS¹ì -í-œ				with fairly static business rules. This architecture is
ë...¼ë¦¬ë¥¼				less suited for dispersed, heterogeneous
³ ì'용				environments with rapidly changing rules. As such,
�˜				relatively few IS organizations are using two-tier
ì (TM) €				client/server architectures to provide
�€ 사이				cross-departmental or cross-platform enterprise-wide
Ã¬-­ì„ ì				solutions Since the bulk of application logic exists on
�¤.				the PC client, the two-tier architecture faces a
�´Ã¬-¸íS¸				number of potential version control and application
ê³¼ì -은				re-distribution problems. A change in business rules
또í-œ				would require a change to the client logic in each
ê°€				application in a corporation's portfolio, which is
ì (TM) € ê° (TM)				affected, by the change. Modified clients would have
이				to be re-distributed through the network - a
				potentially difficult task given the current lack of
�í...Œì´ì...˜				robust PC version control software and problems
CPU 및				associated with upgrading PCs that are turned off or
주변				not "docked" to the network. System security in the
장치ì (TM)				two-tier environment can be complicated since a user
€ 상호				may require a separate password for each SQL
‹¤				server accessed. The proliferation of end-user query
현지				tools can also compromise database server security.
�„				The overwhelming majority of client/server
«Â‹Â¤.				applications developed today are designed without
�´Ã¬-¸íS¸				sophisticated middleware technologies, which offer
�˜				increased security. Instead, end-users are provided a
ì¤'ìš"í-œ				password, which gives them access to a database. In
Â˜ í-œê°œëS"				many cases this same password can be used to
그래í"½				access the database with data-access tools available
				in most commercial PC spreadsheet and database
˜ì´ìS¤				packages. Using such a tool, a user may be able to
(GUI)이다.				access otherwise hidden fields or tables and possibly
일반ì				corrupt data. Client tools and the SQL middleware
ìœ¼ë¡œ				used in two-tier environments are also highly
운영				proprietary and the PC tools market is extremely
�˜				volatile. The client/server tools market seems to be
�€ i.e ì°½				changing at an increasingly unstable rate. In 1994, the
매니ì				leading client/server tool developer was purchased by
€				a large database firm, raising concern about the
				manufacturer's ability to continue to work
ì-¡ì...˜ì„				cooperatively with RDBMS vendors, which compete
ê³ , ì				with the parent company's products. The number
„ì‹œì-				two-tool maker lost millions and has been labeled as a
창을				takeover target. A firm also in the midst of severe
ªÂ³ ì°½ì-				financial difficulties and management transition supplies
있ëS"				the tool, which has received some of the brightest
¼				accolades in early 1995. This kind of volatility raises
�‹¤.				questions about the long-term viability of any
서버				proprietary tool an organization may commit to. All of
�ìS¤: -				this complicates implementation of two-tier systems -
서버				migration from one proprietary technology to another
�ìS¤				would require a firm to scrap much of its investment
�«ÂžÂ¨)ëS"				in application code since none of this code is portable
ìš"구된				from one tool to the next.
ì-...무를				Three tier:- Most sophisticated Web based
�„Âœ				applications, which involve data entry, are based on a
�´Ã¬-¸íS¸				3 tier client server architecture. The 3 tiers are
				• The Client (Web Browser)
�«Â‹Â¤.				• The Web Server/Application
서버				Server
«ÂžÂ¨Ã¬ÂÂ€				• The Database Server The tree
�´Ã¬-¸íS¸				tier architecture (Figure 3) attempts to overcome
«ÂžÂ¨Ã¬-ì„œ				some of the limitations of the two-tier scheme by
일반ì				separating presentation, processing, and data into
ìœ¼ë¡œ				separate, distinct software entities (tiers). The same
ìš"구를,				types of tools can be used for presentation as were
‹¤				used in a two-tier environment, however these tools
�€Ã«Â²				are now dedicated to handling just the presentation.
이ìS¤				When the presentation client requires calculations or
¼ 받ê³				data access, a call is made to a middle tier
ê°±ì‹				functionality server. This tier can perform calculations
은,				or can make requests as a client to additional servers.
�´Ã¬-¸íS¸				The middle tier servers are typically coded in a highly
ìš"ì²­ì-				portable, non-proprietary language such as C.
�€				Middle-tier functionality servers may be multi-threaded
�±ÃªÂ³Â¼				and can be accessed by multiple clients, even those
파견				from separate applications. Although three-tier
�				systems can be implemented using a variety of
«Â‹Â¤.				technologies, the calling mechanism from client to
ë-Œë-Œë¡œ				server in such as system is most typically the
서버				remote procedure call or RPC. Since the bulk of
«ÂžÂ¨Ã¬ÂÂ€				two-tier implementations involve SQL messaging and
일반 ì				most three-tier systems utilize RPCs, it is reasonable
ì´ê³				to examine the merits of these respective request
복잡í-œ				response mechanisms in a discussion of architectures.
ˆìS¤				RPC calls from presentation client to middle-tier
�„				server provide greater overall system flexibility than
‹¤.				the SQL calls made by clients in the two-tier
서버				architecture. This is because in an RPC, the
근거í-œ				requesting client simply passes parameters needed
ê³¼ì -은				for the request and specifies a data structure to
€Ã«ÂÂ„				accept returned values (if any). Unlike most two-tier
�크ì-				implementations, the three-tier presentation client is
다른				not required to "speak" SQL. As such, the
				organization, names, or even the overall structure of
�". 이				the back-end data can be changed without requiring
서버ëS"				changes to PC-based presentation clients. Since SQL
주인				is no longer required, data can be organized
운영				hierarchically, relationally, or in object format. This
체계				added flexibility can allow a firm to access legacy
또ëS"				data and simplifies the introduction of new database
�크				technologies.
파일				Figure 3 - Three-Tier Architecture. Functionality
Â¼ 수				servers handle most of the logic processing.
�;				Middle-tier code can be accessed and utilized by
서버ëS"				multiple clients In addition to the openness stated
ê·¸ ë-Œ ì				above, several other advantages are presented by
œê³µí-œ				this architecture. Having separate software entities
ë'˜ 다				can allow for the parallel development of individual
파일				tiers by application specialists. It should be noted that
시ìS¤í...œ				the skill sets required to develop c/s applications
¤ 및 ì‹				differ significantly from those needed to develop
ì²­				mainframe-based character systems. As examples,
¤Ã¬ÂÂ´Ã«Â‹Â¤.				user interface creation requires an appreciation for
또ëS" ì-´ë-				platform and corporate UI standards and database
í-œ				design requires a commitment to and understanding
�«S", 다른				of the enterprise's data model. Having experts focus
�©				on each of these three layers can increase the
기계ëS"				overall quality of the final application. The three-tier
ì‹ ì²­				architecture also provides for more flexible resource
¤Ã«Â¥Â¼ ì				allocation. Middle-tier functionality servers are highly
�¤. 서버				portable and can be dynamically allocated and shifted
�ìS¤ëS"				as the needs of the organization change. Network
�€Ã«Â²				traffic can potentially be reduced by having
이ìS¤				functionality servers strip data to the precise
인쇄				structure required before distributing it to individual
기계,				clients at the LAN level. Multiple server requests and
통ì‹				complex data access can emanate from the middle
ì-°ê²°,				tier instead of the client, further decreasing traffic.
또ëS"				Also, since PC clients are now dedicated to just
ë†'은				presentation, memory and disk storage requirements
ê°-í (TM) "í-˜				for PCs will potentially be reduced. Modularly designed
žì (TM) € ê°				middle tier code modules can be re-used by several
(TM) 은				applications. Reusable logic can reduce subsequent
ê³µìœ				development efforts, minimize the maintenance
�„				workload, and decrease migration costs when
«S"				switching client applications. In addition, implementation
�›¨ì-´				platforms for three tier systems such as OSF/DCE
«Â¡Âœ ìž'ë (TM)				offer a variety of additional features to support
í-œë‹¤.				distributed application development. These include
서버				integrated security, directory and naming services,
�ìS¤ëS" ìœ				server monitoring and boot capabilities for supporting
사í-œ ì‹				dynamic fault-tolerance, and distributed time
ì²­ì-				management for synchronizing systems across
일반ì				networks and separate time zones. There are of
ì¸				course drawbacks associated with a three-tier
후부				architecture. Current tools are relatively immature and
ì-...무를				require more complex 3GLs for middle tier server
‹¤.				generation. Many tools have under-developed facilities
�ëS"				for maintaining server libraries - a potential obstacle
�´Ã¬-¸íS¸				for simplifying maintenance and promoting code
서버: -				re-use throughout an IS organization. More code in
• 단				more places also increases the likelihood that a
í-˜ë‚˜				system failure will effect an application so detailed
�´Ã¬-¸íS¸, 단				planning with an emphasis on the reduction/elimination
í-˜ë‚˜				of critical-paths is essential. Three tiers brings with it
서버				an increased need for network traffic management,
•				server load balancing, and fault tolerance. For
다수				technically strong IS organizations servicing customers
�´Ã¬-¸íS¸, 단				with rapidly changing environments, three tier
í-˜ë‚˜				architectures can provide significant long-term gains
서버				via increased responsiveness to business climate
�´Ã¬-¸íS¸				changes, code reuse, maintainability, and ease of
서버				migration to new server platforms and development
�´Ã¬-¸íS¸				environments.
서버				Comparing two and three tire development efforts:-
�´				The graphs in Figures 4-6 illustrate the time to
모ë"ˆë³„				deployment for two tiers vs. three tier environments.
«ÂžÂ¨Ã¬ÂÂ˜				Time to deployment is forecast in overall systems
ë...¼ë¦¬ì				delivery time, not man-hours. According to a Deloitte
ì¸				& Touche study, rapid application development time
서버				is cited as one of the primary reasons firms chose to
�ìS¤ì-ì„œ				migrate to client/server architecture. As such,
¤Ã«Â¥Â¼				strategic planning and platform decisions require an
				understanding how development time relates to
�´Ã¬-¸íS¸				architecture and how development time changes as
ê³¼ì -을				an IS organization gains experience in c/s.
ì»´í"¨í„°				Figure 4 - Initial Development Effort Figure 4 shows
ê±´ì¶-ìˆ				the initial development effort forecast to create
은				comparable distributed applications using the common
이다.				two tier and three tier approaches discussed above.
모ë"ˆë³„				The three tier application takes much longer to
«ÂžÂ¨Ã¬ÂÂ€				develop - this is due primarily to the complexity
˜ 구성				involved in coding the bulk of the application logic in a
부분				lower-level 3GL such as C and the difficulties
¡Âœ				associated with coordinating multiple independent
�›¨ì-´ì˜				software modules on disparate platforms. In contrast,
큰				the two-tier scheme allows the bulk of the application
˜				logic to be developed in a higher-level language within
�€ ë"				the same tool used to create the user interface.
쉬운				Figure 5 - Subsequent Development Efforts
발달				Subsequent development efforts may see three-tier
및 ë"				applications deployed with greater speed than two
나은				tier systems (Figure 5). This is entirely due to the
ìœ				amount of middle-tier code, which can be re-used
�성을				from previous applications. The speed advantage
위í-œ				favoring the three-tier architecture will only result if
±Ã¬ÂÂ„				the three-tier application is able to use a sizable
�£Â„´ë‹¤ëS"				portion of existing logic. Experience indicates that
것을				these savings can be significant, particularly in
˜ 기본ì				organizations, which require separate but closely
ì¸				related applications for various business units. Re-use
가ì				is also high for organizations with a strong enterprise
-으로				data model because data-access code can be written
있다.				once and re-used whenever similar access needs
�´Ã¬-¸íS¸				arise across multiple applications. The degree of
서버				development time reduction on subsequent efforts
�€ ê·¸				will grow as an organization deploys more c/s
�€ ë (TM)				applications and develops a significant library of
일í-œ				re-usable, middle-tier application logic.
기ì-µ				Figure 6 - Client Tool Migration Figure 6 makes the
장소				important case for code savings when migrating from
공간				one client development tool to another. It was
ì-ˆì-				stated earlier that client tools are highly proprietary
모ë'ë¥¼				and code is not portable between the major vendor
ㄴ다ëS"				packages. The point was also made that the PC tools
것을				market is highly volatile with vendor shakeouts and
í-„ìš"로				technical "leapfrogging" commonplace. In a two-tier
í-˜ì§€				environment, IS organizations wishing to move from
ì-SëS"다ëS"				one PC-based client development platform to another
것을				will have to scrap their previous investment in
�¬Â„Âœ				application logic since most of this logic is written in
게				the language of the proprietary tool. In the three-tier
¼ 멀리				environment this logic is written in a re-usable middle
ì·¨í-œë‹¤.				tier, thus when migrating to the new tool, the
이				developer simply has to create the presentation and
ê±´ì¶-ìˆ				add RPC calls to the functionality layer. Flexibility in
로,				re-using existing middle-tier code can also assist
부르ëS"				organizations developing applications for various PC
단위ëS"				client operating system platforms. Until recently there
´ì-¸íS¸"ì-				were very few cross-platform client tool
�¤Ã«Â¥Â¼				development environments and most of today's
ì €ê²ƒ)				cross-platform solutions are not considered
�³ , 불린				"best-of-breed". In a three-tier environment separate
단위ëS"				client tools on separate platforms can access the
°Â€ 된다				middle tier functionality layer. Coding application logic
�¤Ã«Â¥Â¼ ì				once in an accessible middle tier decreases the overall
œê³µí-˜ëS")				development time on the cross-platform solution and
ì €ê²ƒ.				it provides the organization greater flexibility in
�˜ ë...¼ë¦¬ì				choosing the best tool on any given platform.
ì¸				The characteristics of client/server architecture:-
�				The basic characteristics of client/server architectures
ê·¸ë"¤ì˜				are:
위í-œ ì				1) Combination of a client or front-end portion that
í-©í-œ				interacts with the user, and a server or back-end
-´ì (TM) €				portion that interacts with the shared resource. The
�›¨ì-´				client process contains solution-specific logic and
í"Œëž˜íS¸í				provides the interface between the user and the
(TM) ˆì-				rest of the application system. The server process
ë‹¬ë ¤				acts as a software engine that manages shared
�´Ã¬-¸íS¸ì (TM) €				resources such as databases, printers, modems, or
¥Â¼				high-powered processors.
ëS"이다.				2) The front-end task and back-end task have
예를				fundamentally different requirements for computing
ë"¤ë©´,				resources such as processor speeds, memory, disk
처리				speeds and capacities, and input/output devices.
„ 위í-œ				3) The environment is typically heterogeneous and
íS¹ë³„í-œ				multiFinder. The hardware platform and operating
œÂ¼Ã«Â¡Âœ				system of client and server are not usually the same.
질문,				Client and server processes communicate through a
또ëS"				well-defined set of standard application program
í"Œëž˜íS¸í				interfaces (API's) and RPC's.
(TM) ˆì-				4) An important characteristic of client-server
달리ëS"				systems is scalability. They can be scaled horizontally
파일				or vertically. Horizontal scaling means adding or
¥Â¼				removing client workstations with only a slight
¸°				performance impact. Vertical scaling means migrating
위í-˜ì-¬				to a larger and faster server machine or multiservers.
²Œ				Client-server applications:-
�˜ê³				We define a client-server application as:
�ëS"				An application system in which logically separate
í"Œëž˜íS¸í				software components are integrated together via
(TM) ˆì-				client-server relationships. In a client-server
달리ëS"				relationship, one part of an application (the client end)
�€Ã«Â²				uses a service provided by the other part (the
이ìS¤				server end). The latter is often a shared resource,
관리				used by many clients. Although integrated together
체계				via the client-server relationship, the parts remain
서버.				separate. We refer to them as being logically
�크				separate because they need not be physically
계산				remote from one another (they might be in the
ê±´ì¶-ìˆ : -				same computer). We describe client-server application
Oracle의				software here in three steps: splitting an application,
�크				joining separate applications together, and distributed
계산				application structure
ê±´ì¶-ìˆ				Splitting an application:- Figure 5 Application software
(NCA)ëS"				modularity There are many ways of partitioning
3개의				application software into separate components.
개ë...ì-				However, the content of most applications can
의í-´ ë¶ (TM)				usually be classified under three different technical
잡을				headings: data management, application logic and
수				presentation. This is illustrated in figure 5. If the
있다:				application is to be split into two parts (one part on a
1.The ì›"ë"œ ì (TM)				client platform, the other on a server platform), the
€ì´ë"œ				split can be made at either of the two boundaries
웹은 ì				between functions, or inside one of the three
-말				functions. Consequently there are five main ways of
�«S"				splitting a centralized or personal application into two
¤Ã¬ÂÂ´Ã«Â‹Â¤.				parts between which there is a client-server
2.The Java				relationship. This is the basis of the popular
가상				classification into five client-server styles, which is
머ì‹				promoted by the Gartner Group. It is illustrated in
은 (또ëS"				figure 6.
곧 ë				Figure 6 Five generic styles of basic client-server
�¤) 웹				structure
�°Ã¬ €				The details need not concern us here. The important
ì-ˆì-ì„œ				point is that different styles suit different needs and
묻힌 ì				circumstances:
-말				• The two styles on the left of
�«S"				the diagram are typical of centralized interactive
¤Ã¬ÂÂ´Ã«Â‹Â¤.				applications that have been adapted to client-server
Oracle				by means of graphical interface technology, terminal
�€Ã«Â²				emulation, etc.
이ìS¤				• The style in the middle of the
ì-"진을				diagram is typical of object-oriented distributed
가진				applications and distributed TP applications in which
ì‹ ì²­				data and function are encapsulated together behind
납í'ˆ (ìœ				application interfaces
닉ìS¤				• The two styles on the right of
큰				the diagram are typical of data-centered applications
),				using client-server 4GL development tools and
�´Ã¬-¸íS¸ì-				relational database products Some applications
경량 Java ì-				combine all three areas of function (presentation,
이ì...˜,				application logic and data management) at the
및 ì¤'ì- (TM)				personal platform. Also, different styles may occur in
층을				combination at the same platform.
위í-œ 3.A 3				Joining applications together:- One of the great
층층				strengths of client-server is the ability to join
2 사이				separate applications together. This can be done in
를 ì				many ways; but upon the principles used in 2.4.1,
° 위í-˜ì-¬				there are essentially three levels at which applications
¥Â¼				can interface with one another. This is illustrated in
�‹¤".				figure 7.
OracleëS"				Figure 7 Three levels at which applications can be
1998ë...„은				joined together
1ì›"ì-ì„œ				The main characteristics and advantages and
릴리ìS¤				disadvantages of these three approaches are:
10.7 NCA (웹				• At presentation level:
œ ì‹				Interaction at this level is achieved via direct data
청)를				exchange (DDE) within a window management
«S"				system, or via scripting; see [Duxbury, 1994], in which
¦. 2000ëS"				software uses an application's user interface by
�‹Â¤				simulating a human user. This kind of technique is
방출 10.7				often referred to as screen scraping. It is very useful
NCA로, OracleëS"				for accessing legacy applications, but leads to
ë˜'ë˜'í-œ				software maintenance problems if the user interfaces
„Âœ				need to change.
˜ ì-´ë				• At application function level:
¤ì›€ì-				Interaction at this level is in terms of business
ê³ ê°				functions. Therefore, the inter-application requests
-				are about the business meanings of the application
�‹Â¤.				(and not its presentation or database encoding). This
기ëS¥ì´				has the advantage of keeping their internal designs
10 SCì (TM) € 10 NCA				separate from their external interactions. There are
�서 ë (TM)				fewer software maintenance problems.
일				• At data management level:
�도, OracleëS"				Interaction at this level is by direct access to the
� 깁ëS" ì				other application's database. This is common practice,
„ëžµì-				but leads to software maintenance problems when
웹				application data structures change.
배치í-œ				The first and third approaches inhibit potential for
¬-ì„œ				change, the second does not. Further distinctions can
돌ë				be made between direct and indirect interaction
�‹Â¤. 이 ì				between applications, synchronous and asynchronous
„략은				interaction, and externally programmed interaction
또í-œ				and internally programmed interaction.
ë³´ì¡´				Distributed application structure:- Distributed
식í'ˆ				applications are evolving towards richly connected
주문í (TM)				network structures of the kind illustrated in figure 8.
"를				The circles represent separate software
„다.				components, and the lines represent client-server
í-ê²S				relationships between them. This is typical of the kind
„ ì				of structure that results from use of object-oriented
�° 후ì-				design and distributed object management.
´ 지금				Figure 8 Complex distributed application There is also
ë"				large-scale structure of distributed application
쉽기				systems (within which the individual client-server
모ì-'				relationships occur). Typically, three tiers of application
ê¸°ìˆ ì´				software can be discerned in the large-scale
�				structure:
�기 웹				• Front tier: Application software
ªÂ¸Â°				(and databases) at personal platforms, providing all
¬- 있ëS"				kinds of application services, using local resources and
				remote resources. Typically, the platforms are PCs.
달리ê³				This tier is where the greatest amount of computer
기				power and of new application software is now being
ë-Œë¬¸ì-.				deployed.
깁ëS" ì				• Middle tier: Application software
„ëžµì-				(and databases) at server platforms, providing the
있ëS"				back-end of personal applications, shared workgroup
다름				services and task-oriented services. Typically, the
ë-Œë¬¸ì-,				platforms are UNIX or PC. This tier provides rapid
OracleëS" ê³ ê°				adaptation to business process change, without
사용				needing changes to the back tier. It puts boundaries
ë˜'ë˜'í-œ				around the turbulence and uncertainty generated in
�´Ã¬-¸íS¸ë¥¼				the volatile world at the first tier, where all the users
ì-„ë‹ (TM)				are. It also provides lateral linkage across the
니다				enterprise (e.g. electronic mail services).
�³ ë (TM)				• Back tier: Application software
일í-œ				and databases at server platforms providing
경우ì-				corporate information services. These are usually
있ëS" 10.7				functionally partitioned (e.g. accounts, manufacturing,
NCA를 í'€ì-´				personnel). Typically, the platforms are mainframes.
ë†"ëS"다.				This tier provides the core of shared and long-lived
OracleëS" 그런				information assets that everything else depends on.
„				There are strong guarantees of data integrity, and
�지				the applications and databases are stable, and their
ì-S을				design changes rather slowly.
�¤. Customerncharactermode				This structure separates different kinds of concerns,
€ 웹				which used to be bundled together in centralized
배치í-œ				computingchapter-8
¬- 직ì 'ì				Important of client server:-
ìœ¼ë¡œ				Advantages of Client-Server:-
ì´ë (TM)				ØPotential of reduced cost
í-´ì-¼				ØImproved performance
í-œë‹¤				ØIncreased security
장 5				ØMore GUI application
�´Ã¬-¸íS¸				ØGives people the opportunity to make
서버				change for better
ê¸°ìˆ : -				ØBetter SW development tools once
�´Ã¬-¸íS¸				established
서버				ØExploits existing H/W, S/W
기ìˆ				configurations
은 잘				ØMatches distributed business models
�� 4개의				ØScalable
�„œ				ØFlexibility and cost saving
„ í† ë¡				ØFlexibility business modeling
í-˜ëS"				ØMaximum technology component
경우ì-				choice
�«Â‹Â¤:				ØEfficient use of computing resources
1.Personal				ØData interchangeability and
í"Œëž˜íS¸í				interoperatability
(TM) ˆ				ØEnhanced data sharing
2.Server				ØIntegrated services
í"Œëž˜íS¸í				ØSharing resources among devices
(TM) ˆ				platforms
3.Client 서버				ØLocation independence data and
�¬-´				process
4.Client 서버				ØTechnology revolution
공구ì (TM)				ØFuture technology
€				ØRelational databases
¤				ØDisadvantages of Client-Server:-
이				ØHeavy up-front cost
				ØInitial performance decline
�€ ê·¸ë"¤				ØLack of skilled professionals
사이				ØNeed of rewrite a lot of software
©ì´				ØNeed for retraining user
있을				ØDependability- when the server goes
수				down, operational cases
�도, íS¹ìœ				ØLack of mature tools
í-˜ë‹¤.				ØLack of the scalability-network
기간				operating system (e.g. novel NetWare, window NT
í"Œëž˜íS¸í				server) are not very scalable
(TM) ˆì€				ØHigher then anticipated costs
-´ì (TM) €				ØHarder to build
운영				ØLess stable
체계				ØSusceptible to network load
�›¨ì-´ì˜				ØLacking in the specialists
ì (TM) „ì „í-œ				ØDifficult to debug
ì¡°í-©ì¸				ØDifficult to test
ì»´í"¨í„°				Client/Server Business Application Architectures:
í"Œëž˜íS¸í				Traditional applications architectures have been based
(TM) ˆì„				on function today, to meet the needs of the
"				business an application architecture should reflect the
�„Âœ				complete range of business requirements.
œë‹¤.				Therefore, client/server computing demands a three
개인				layer view of theorganization.
í"Œëž˜íS¸í				1 The user interface layer, which implements the
(TM) ˆ: -				functional model
개인				2 The business function layer, which implements the
í"Œëž˜íS¸í				process model
(TM) ˆì€				3 The data layer, which implements the information
ì-„마				model It should be noted that this application
�´Ã¬-¸íS¸				architecture does not demand multiple hardware
서버				platforms, although such technology can be utilised,if
ê¸°ìˆ ì˜				the environment is robust and reliable enough and the
가장				business is prepared to pay the additional costs
íS¹ìœ í-œ				associated with workstation and LAN technology.
다.				Business Benefits: - There is a perceived need for
우리ëS"				vendor independence. This includes application
개인				development methodologies, programming paradigms,
í"Œëž˜íS¸í				products and architectures. - Organization have
(TM) ˆì„				changed from steep hierarchies to flattened
�¼ ê° (TM) 음				hierarchies - Network management is replacing
ì				vertical management - There is a change to team
�:				based management - The customer should have a
�크ì-				single point of contact for all business with the
�«S",				organization - The customer should deal with the
ì»´í"¨í„°				same person over multiple contacts. - The user will
í"Œëž˜íS¸í				perform as much processing as possible during
(TM) ˆì€				customer contact time - The time required to
œ				complete the work will be minimized - There is a
직관ì				need for empowerment of staff and audit trail of
ì¸				actions - Multi-skilled and multi-function teams need
				access to multiple applications
˜ì´ìS¤				Different types of servers:- The simplest form of
및				servers are disk servers and file servers. With a file
기ì-...의				server, the client passes requests for files or file
대ì‹				records over a network to the file server. This form
으로				of data service requires large bandwidth and can
ì-...무를				slow a network with many users down considerably.
�기				Traditional LAN computing allows users to share
위í-˜ì-¬				resources, such as data files and peripheral devices,
개인				by moving them from standalone PCUs onto a
				Networked File Server (NFS). The more advanced
�ªÂ¸Â° ì				form of servers are database servers, transaction
�¤. 이				server and application servers (Orfali and Harkey
�� 숫자				1992). In database servers, clients pass SQL
2ì-ì„œ				(Structured Query Language) requests as messages
�다.				to the server and the results of the query are
개인				returned over the network. The code that processes
í"Œëž˜íS¸í				the SQL request and the data resides on the server
(TM) ˆì€				allowing it to use its own processing power to find
˜ 싸ê³				the requested data, rather than pass all the records
게 ê°-ë				back to a client and let it find its own
¥í-˜,				Data as was the case for the file server. In
�Ã¬ÂÂ˜				transaction servers, clients invoke remote procedures
ë„"은				that reside on servers, which also contain an SQL
ì„				database engine. There are procedural statements on
택이				the server to execute a group of SQL statements
있다.				(transactions), which either all succeed or fail as a
�¬ÂÂ˜				unit. The applications based on transaction servers
많은				are called On-line Transaction Processing (OLTP) and
다른				tend to be mission-critical applications, which require
ì¢...류ëS"				1-3 second response time, 100% of the time and
개인				require tight controls over the security and integrity
í"Œëž˜íS¸í				of the database. The communication overhead in this
(TM) ˆ				approach is kept to a minimum as the exchange
ë©´ MS/DOS PC,				typically consists of a single request/reply (as
ì°½ PC, OS/2 PC, ìœ				opposed to multiple SQL statements in database
닉ìS¤				servers). Application servers are not necessarily
�í...Œì´ì...˜,				database centered but are used to server user
Apple				needs, such as.
매킨í†				Download capabilities from Dow Jones or regulating a
시, 및				electronic mail process. Basing resources on a server
각ì¢...				allows users to share data, while security and
소í˜-				management services, which are also based in the
¼ 수				server, ensure data integrity and security.
있다;				Special types of Architecture: IBM's System
˜ 일반ì				Application Architecture: SAA is a collection of
ì¸				selected software interfaces, conventions, and
�ëS" 오ëS˜				protocols that are used as a framework for
Microsoft Windows				developing consistent, integrated applications across
운영				the major IBM computing environments.
�¼ 가진 IBM				Four major components of this architecture are: -
호í (TM)				Common User Access (CUA) defines conventions for
˜ìš©				GUI look and feel. - Common Programming Interface
PC이다.				(CPI) provides languages, tools, and APIs that give
그런				applications greater portability and more consistent
í"Œëž˜íS¸í				user interfaces across multiple platforms. - Common
(TM) ˆì€				Communication Support (CCS) supports existing
지금				communications standards, such as LU 6.2. - Common
�¤ ì-´ë""ë"				Applications, written by IBM, will serve as
지				demonstrations of SAA concepts and make it easy
보편ì				for users to migrate between systems.