On Wed, 26 Feb 2020 at 17:21, Eric Van Tol <e...@atlantech.net> wrote:
> I could be wrong, but the way I understand it, the original SCB only has > 80Gb/s of fabric capacity. While the 16XGE MPC will technically work, it will > only work at half capacity. For full throughput capability, you would need > two SCBs for a total of 160Gb/s fabric capacity, but if you lose one of the > SCBs, you lose half your capacity. Again, my understanding could be wrong, so > someone please correct me if so. I believe you're wrong. SCB is 3.125Gbps SERDES, MX240/MX480 are (8+8) + (8+8) per MQ to/from fabric and MX960 is 8+8+8. So both SCB's up is 100Gbps on MX240/MX480 serdes per MQ to/from fabric, one down is 50Gbps (40Gbps needed for ucast) All SCB up is 75Gbps on MX960 serdes per MQ, one down is 50Gbps. 16X10GE unary replicates, without fabric failures SCB has enough on MX240/MX480 and nearly enough on MX960. With Fabric failure, not so great. MPC[12] binary replicates (so it needs double SERDES, 80G serdes to support 40G traffic). Has enough for unicast with fabric failures, but not for multicast. -- ++ytti _______________________________________________ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp
