光纤协议仲裁环ArbitratedLoop.ppt

* Device 1 Device 2 Device 4 Device 3 ARBITRATED LOOP Continual Idle frame is sent around the loop Idle frame Device 1 Device 2 Device 4 Device 3 ARBITRATED LOOP Device 3 Idle frame Device 2 Idle frame Device 4 Idle frame Device 1 Idle frame Each device replicates the idle frame and inserts its own signature (AL_PA) Device 1 Device 2 Device 4 Device 3 ARBITRATED LOOP Device 1 arbitration When a device; device 1, wishes to send data around the loop the device must first send out an Arbitration frame, a request/intension to communicate Device 1 arbitration Device 1 arbitration Device 1 arbitration Device 1 Device 2 Device 4 Device 3 ARBITRATED LOOP If device 1 gets it’s arbitration frame back it means that no other device is communicating so device 1 can then communicate Device 1 Data to device 3 Device 1 Data to device 3 Device 1 blocks all idle and arbitration frames from other device whilst it sends data Device 1 Device 2 Device 4 Device 3 ARBITRATED LOOP To prevent a device monopolising the loop an arbitration sequence is limited to 32 frames. Idle frame If device 1 wanted to communicate further it must first send and receive an idle frame EXP PRI 14 13 12 3 2 1 0 DAE2 Shunt EXP PRI 14 13 12 3 2 1 0 DAE2 CX Series Architecture EXP PRIor MID 14 13 12 3 2 1 0 DAE2(OS)orDPE SP Shunt Shunt DAE2 Port Bypass Circuits RP/RT RP/RT Port Bypass Circuits RP/RT RP/RT Empty slot ATA DAE2 LCC-ATA Logic …. ATA LCC 1 LCC-ATA Logic …. ATA LCC 2 FC/ATA Translation FC/ATA Translation Switching Logic DAE0 Switching Logic DAE1 DAE3P - Accessing a Drive in DAE0 Pri Port Exp Port Exp Port Pri Port Switching Logic DAE0 Switching Logic DAE1 DAE3P - Accessing a Drive in DAE1 Pri Port Exp Port Exp Port Pri Port For CX-series arrays it is much simpler. Into the first enclosure – drives 0 to 14, then to the next enclosure – drive 0 to 14, so on and so forth until the last drive on the last enclosure. The return path is through the cables and LCCs to get to the SP. In the DA