Hardware
Hope This Helps.
LVM
※ Raw device 형태의 LV(Logical volume)은 백업되지 않는다.
0) 테스트 환경
1) SMIT : VG 백업
#smitty savevg
2) 백업 확인
#smitty lsbackvg
1) 백업 복구
#smitty restvg
2) 복구 확인
1) 노드 구성 계획
2) Inter-Connect(Heartbeat) Interface
3) GPFS 공유파일시스템(MMFS)의 용량과 각 NSD disk(LUN)은 몇개로 가져갈 것인지?
4) Remote GPFS Cluster를 사용할 것인지?
5) Communication protocol(통신방식)은 어떤 것을 사용할 것인지?
1) /etc/hosts에 상호 노드 등록
2) 각 노드에서 공개키 교환
3) 디스크 속성
1) 설치 (모든 노드에서 GPFS 설치)
2) 패스 등록(모든 노드에서 패스등록)
PATH=$PATH:/usr/bin:/usr/dt/bin:/etc:/usr/ucb:$HOME/bin:/usr/bin/X11:/sbin:/usr/lpp/mmfs/bin:.
1) 노드 정의
2) NSD 정의
3) GPFS Cluster 생성
4) GPFS 라이센스 동의 및 노드 상태 확인
5) GPFS Startup 및 노드 상태 확인
6) NSD 및 파일시스템 생성
7) Tie breakers disk 지정
8) GPFS 파일시스템 마운트
Figure1) RAID 0 동작원리
1) 동작 방식
2) 특징
Figure2) RAID 1 동작방식
Figure3) RAID1+0 동작방식
1) 동작방식
Figure4) RAID5 동장방식
Figure5) RAID6 동작방식
RAID0
RAID1
RAID1+0
RAID5
RAID6
사용 가능한 디스크 (N개 기준)
N개
N/2개
N-1개
N-2개
최소 드라이브 개수
2개
4개
3개
최대 용량
디스크의 수 * 디스크의 용량
(디스크의 수 / 2) * 디스크의 용량
디스크의 수 / (RAID 1로 묶는 디스크 개수) * 디스크 용량
(디스크의 수 -1) * 디스크의 용량
(디스크의 수 – 2) * 디스크의 용량
장점
•빠른 I/O 성능
•구성한 디스크 개수만큼 용량 모두를 활용 가능
•높은 안정성
•읽기 성능에서 비교적 우수한 성능
•Striping의 이점과 mirroring의 이점을 동시에 가짐
•디스크 장애 시 복구 수월
•패리티 연산의 부하를 줄임
•패리티를 통한 안정성과 Striping을 통한 성능 우수
•하위 level 대비 안정성 증가
•두 개의 드라이브 장애까지 허용
단점
•낮은 안정성
•절반의 디스크밖에 쓰지 못하므로 비용이 높음
•기술적으로 구현 난이도가 높음
•두 개 이상의 디스크가 손상될 시 데이터 손실
•하위 level 대비 비용 증가