: Session 3. Designing and Implementing Compaq StorageWorks Module 1.

강사 : 강북팀장

Session 3

Designing and ImplementingCompaq StorageWorks

Module 1

목 차1. Storage System 구성요소에 대한 이해

2. Storage Subsystem 관리 환경 설정 및 구성

3. HSG80 Controller Configuration 방법

4. Raid 종류와 Setting 방법

5. Troubleshooting

4

RA8000 SCSI Buses 구조 All devices in the same

column are on the same SCSI bus, or port.

All devices in the same row have the same device address, or target ID.

Device addresses 4 and 5 are reserved and not to be used.

SBB Power Supply Placement

SBB Power Supplies

• At least four are necessary to power the cabinet

- Five power supplies are required for a “standard” power configuration (4+1)

- One Power Supply is “redundant” and will take over for any single failed power supply.

- Eight power supplies and a “B” power entry module are required

- for fully redundant power (4+4 configuration)- Power supplies are “hot-swappable” (if there are

more than four)

7

Environmental Monitoring Unit(EMU)

EMU Error Detection

– The EMU will detect

• High temperature• Failed disk drives• Cache battery low voltage or failure• Controller errors• Ultra SCSI termination errors• Blower failures

EMU and System Shutdown

– The EMU will shutdown the subsystem cabinet if:

• A blower is removed for more than eight minutes• If all four blowers in one vertical bank are defective and

not replaced within 8 minutes• If there are less than four functioning power supplies• If any two temperature sensors register 50 Degrees

Centigrade(122 Degrees F).

This is the EMU default setting.It can be changed.

Power Verification and Address Unit (PVA)

1. Power switch2. UPS Communications3. SCSI target wheel

Power Verification and Address Unit (PVA)

– Works with the EMU and acts as a redundant fault monitor

– The PVA will shut down the subsystem if the EMU is removed for more then eight minutes

– The PVA cannot be hot swapped– Valid PVA SCSI settings are 0, 2 and 3 only – If the PVA SCSI setting must be changed, the

cabinet must be shut down and powered off for the changes to take effect

12

Setting the ID for the Enclosure

0, 2, and 3 are the only valid PVA settings

13

HSG80 RAID controller

Controller Operator Control Panel (OCP)

1. FC Ports2. ACS Slot3. Maintenance Port4. Operator Control Panel

Array Controller Software

• PCMCIA Card– Each controller requires a PCMCIA card to operate– PCMCIA card contains controller firmware– Firmware is downloaded into volatile memory at startup

• Cache Modules

– Left cache module : top controller

– Right cache module : bottom controller

External Cache Modules

External Cache Batteries (ECB)

• Cache Batteries– Cache batteries are checked every 24 hours– Weak Cache batteries checked every 4 minutes for

10 hours– When fully charged, an ECB can preserve 512 MB of

cache memory for 24 hours, or 64 MB of cache memory for 48 hours

– Cache battery LEDs• Slow flashing = discharging• Rapid flashing = charging• Steady green = fully charged

– Battery cut-off button disconnects battery from cache

Dual Speed Blowers

– Can be hot swapped

– Can be set to run at high speed continually

– If a blower is removed, the subsystem will shut down after eight minutes

I/O Modules

– Six modules with dual pairs of 68-pin VHD connectors.

– One for each internal SCSI bus

– Connect expansion cabinets to the primary cabinet

– Wide-Ultra SCSI-2, single-ended termination.

– Cable length between cabinets is restricted to 1.5 meters

Console 환경 설정하기• Establishing a CLI Session

– Connect the serial cable (DB9 to RJ12) from the maintenance port of the controller to any terminal or to the COM Port of a PC running a terminal emulation program

• Terminal Setting (using Hyper Terminal)– Use these parameters to establish a session

• 9600 bps• 8 data bits• 1 stop bit• No parity• No flow control

– Press the Enter key

Controller Failover Mode Type

Dual-Redundant (Transparent Failover Mode)

Multi-Bus Failover Mode

Configuring Redundant Controllers

Dual–Redundant Mode

SET FAILOVER COPY = THIS orSET FAILOVER COPY = OTHER

Multi-Bus Failover Mode

SET MULTIBUS COPY = THIS orSET MULTIBUS COPY = OTHER

HSG80 Controller Configuration 방법

1. HSG80 Controller Terminal Emulation Program - Settings (using the Hyper Terminal)

2. 설정 확인 - SHOW THIS

3. HSG80 Controller 초기화 ( 미사용컨트롤러 물리적제거후 )- CONFIGURATION RESET- SET NOFAILOVER- Controller 의 후면 확인 (Bottom Controller down 된 상태 확인 )- CONFIGURATION RESET 명령 수행- DOWN 된 BOTTOM CONTROLLER 의 RESET BUTTON 을 누른다 .

4. SHOW THIS FULL (NODE ID 확인 ) - NODE ID 가 모두 0 로 설정되어 있는 것을 확인

5. SET THIS_CONTROLLER NODE_ID = XXXX-XXXX~

- NODE ID 설정 후 Controller 가 Restart 되는 것이 요구됨 .

6. RESTART THIS - Controller 가 restart 된 후에 Controller Failover mode, Time 과

Cache 설정 등이 요구된다는 메시지를 확인할 수 있다 .

7. SHOW THIS - Controller 의 Failover mode 및 Time 등 기타 설정이 되어있지 않은

상태를 확인한다 .

8. SET FAILOVER COPY = THIS (SET MULTIBUS COPY = THIS)

9. SHOW THIS - Time, Mirrored Cache 설정이 되어 있지 않은 상태를 확인

10. SET THIS TIME = 07-NOV-2001:07:15:00 - 날짜 setting 시에 월에 해당되는 항목 설정 시 주의

11. SET THIS MIRROR_CACHE - Controller 가 restart 된다 .

12. SHOW THIS - 결과 확인 후 마친다

13. SET THIS PORT_1_TOPOLOGY = LOOP_SOFT - FC-Hub 일 경우 사용

set THIS PORT_1_TOPOLOGY = FABRIC - FC-Switch 일 경우 사용

Raid 종류와 Setting 방법

1. JBOD2. RAID 0 : Stripe Set3. RAID 1 : Mirror Set4. RAID 5 : Raid Set5. RAID 0+1 : Striped Mirrorset6. Spare Set

• DISK 인식 방법

1. Automatic 방식 - RUN CONFIG

2. Manual 방식 - ADD DISK PTLAddress PORT TARGET LUN (EX : ADD DISK DISK10000 1 0 0)

DISKPTTLL Rule : P = Port Number DISKPTTLL Rule : P = Port Number TT = SCSI ID Number TT = SCSI ID Number LL = Always 00LL = Always 00

HSG80> INITIALIZE DISK10000HSG80> INITIALIZE DISK10000HSG80> ADD UNIT D0 DISK10000HSG80> ADD UNIT D0 DISK10000HSG80> SHOW DISK10000HSG80> SHOW DISK10000HSG80> SHOW D0HSG80> SHOW D0

Step 1 - JBOD

Step 2 - RAID 0 : Stripe SetHSG80> ADD STRIPESET STR1 DISK20000 HSG80> ADD STRIPESET STR1 DISK20000

DISK30000DISK30000HSG80> INITIALIZE STR1HSG80> INITIALIZE STR1HSG80> ADD UNIT D1 DISK10000HSG80> ADD UNIT D1 DISK10000HSG80> SHOW STR1HSG80> SHOW STR1HSG80> SHOW D1HSG80> SHOW D1

HSG80> ADD MIRRORSET MIRR1 DISK40000 HSG80> ADD MIRRORSET MIRR1 DISK40000 DISK50000DISK50000HSG80> INITIALIZE MIRR1HSG80> INITIALIZE MIRR1HSG80> ADD UNIT D2 MIRR1HSG80> ADD UNIT D2 MIRR1HSG80> SHOW MIRR1HSG80> SHOW MIRR1HSG80> SHOW D2HSG80> SHOW D2

Step 3 - RAID 1 : Mirror Set

Step 4 - RAID 5 : Raid Set

HSG80> ADD RAIDSET R1 DISK10100 DISK20100 HSG80> ADD RAIDSET R1 DISK10100 DISK20100 DISK30100DISK30100HSG80> INITIALIZE R1HSG80> INITIALIZE R1HSG80> ADD UNIT D3 R1HSG80> ADD UNIT D3 R1HSG80> SHOW R1HSG80> SHOW R1HSG80> SHOW D3HSG80> SHOW D3

HSG80> ADD MIRRORSET MIRR2 DISK40100 HSG80> ADD MIRRORSET MIRR2 DISK40100 DISK50100DISK50100HSG80> ADD MIRRORSET MIRR3 DISK60000 HSG80> ADD MIRRORSET MIRR3 DISK60000 DISK60100DISK60100HSG80> ADD STRIPESET STR2 MIRR2 MIRR3HSG80> ADD STRIPESET STR2 MIRR2 MIRR3HSG80> INITIALIZE STR2HSG80> INITIALIZE STR2HSG80> ADD UNIT D4 STR2HSG80> ADD UNIT D4 STR2HSG80> SHOW STR2HSG80> SHOW STR2HSG80> SHOW D4HSG80> SHOW D4

Step 5 - RAID 0+1 : Striped Mirrorset

Step 6 – Spare Set

HSG80> ADD SPARESET DISK60200HSG80> ADD SPARESET DISK60200HSG80> SHOW SPARESETHSG80> SHOW SPARESETHSG80> SHOW STORAGEHSG80> SHOW STORAGE

HSG80> SET FAILEDSET AUTOSPAREHSG80> SET FAILEDSET AUTOSPAREHSG80> SHOW FAILEDSETHSG80> SHOW FAILEDSETHSG80> SET FAILEDSET NOAUTOSPAREHSG80> SET FAILEDSET NOAUTOSPAREHSG80> SHOW FAILEDSETHSG80> SHOW FAILEDSET

Failedset Disk Auto Rebuilding 설정

UNIT, STORAGE, DISK 삭제

UNIT : HSG80> DELETE UNITUNIT : HSG80> DELETE UNIT 명 명 HSG80> SHOW UNITHSG80> SHOW UNITSTORAGE : HSG80> DELETE STORAGESTORAGE : HSG80> DELETE STORAGE 명명

HSG80> SHOW STORAGEHSG80> SHOW STORAGEDISK : HSG80> DELETE DISKDISK : HSG80> DELETE DISK 명명

HSG80> SHOW DISKHSG80> SHOW DISK

FAIL DISK REPLACE( 디스크교체시 )

HSG80> SHOW STORAGEHSG80> SHOW STORAGEHSG80> DEL FAILEDSET DISKHSG80> DEL FAILEDSET DISK 명명HSG80> SHOW DISKHSG80> SHOW DISKHSG80> DELETE DISKHSG80> DELETE DISK 명명 ( NEW DISK REPLACE )( NEW DISK REPLACE )HSG80> ADD DISK DISKHSG80> ADD DISK DISK 명명HSG80> SHOW FAILED_RAIDSETHSG80> SHOW FAILED_RAIDSET 명 명 HSG80> SET FAILED_RAIDSETHSG80> SET FAILED_RAIDSET 명 명 NOPOLICYNOPOLICYHSG80> SET FAILED_RAIDSETHSG80> SET FAILED_RAIDSET 명 명 RECONSTRUCT = FASTRECONSTRUCT = FAST ( DATA ACCESS SPEED = LOW( DATA ACCESS SPEED = LOW DATA REBUILDINF SPEED = FASTDATA REBUILDINF SPEED = FAST RAIDSET ONLY )RAIDSET ONLY )HSG80> SET FAILED_RAIDSETHSG80> SET FAILED_RAIDSET 명 명 REPLACE = NEWDISKREPLACE = NEWDISK명명HSG80> SHOW FAILED_RAIDSETHSG80> SHOW FAILED_RAIDSET 명 명 HSG80> SET FAILED_RAIDSETHSG80> SET FAILED_RAIDSET 명 명 RECONSTRUCT = RECONSTRUCT = NORMALNORMALHSG80> SET FAILED_RAIDSETHSG80> SET FAILED_RAIDSET 명 명 POLICY = POLICY = BEST_PERFORMANCEBEST_PERFORMANCE

ADD UNIT {LUN} {Storage Set Name}ADD UNIT {LUN} {Storage Set Name}

•Port 1 displays units D0 through D99Port 1 displays units D0 through D99•Port 2 displays units D100 through D199Port 2 displays units D100 through D199

Example : ADD UNIT D101 M1Example : ADD UNIT D101 M1

Assigning LUNs (Fibre Channel)

Any Server, Anytime, Anywhere!