Overview

When sourcing from a custom fiber optic switch manufacturer, network architects need verified specs on ASIC forwarding capacity, transceiver interoperability, and hardware-level redundancy. This FAQ addresses both pre-sales validation (throughput, backplane) and post-sales troubleshooting (CLI, stacking, power failover) to ensure carrier-grade deployment.

Frequently Asked Questions

Q1: What is the maximum switching throughput and backplane capacity of your custom fiber optic switch?

The maximum switching capacity for a 48-port 10G custom fiber optic switch is 1.44 Tbps (full duplex) with a forwarding rate of 1,071 Mpps for 64-byte packets. Backplane capacity refers to the internal fabric bandwidth; on our custom switches, it matches the switching capacity—1.44 Tbps—eliminating blocking. For 100G uplink models, backplane scales to 6.4 Tbps. Always verify that the backplane equals the sum of all port bandwidths to guarantee line-rate performance.

Q2: Which optical modules (SFP, SFP+, QSFP28) are compatible with your custom fiber optic switch?

Our custom fiber optic switches support standard MSA-compliant transceivers (SFP, SFP+, QSFP28) for 1G, 10G, 25G, 40G, and 100G connections. This includes third-party optics from Finisar, Lumentum, and II-VI after programming verification. For lock-in prevention, our firmware does not enforce vendor-specific coding. However, we provide a validated compatibility matrix: SR/LRM for multimode up to 300m, LR/LW for single-mode up to 10km, and ER/ZR for 40-80km. For custom coding (e.g., Cisco or Arista IDs), add the option at order placement.

Q3: How does this custom fiber optic switch compare to previous generation (e.g., 1G to 10G/40G models)?

The key generational improvements are: (1) Packet buffer: increased from 9MB to 32MB per ASIC, reducing micro-burst drops. (2) Forwarding table scale: MAC address table grew from 32K to 288K, IPv4 routes from 16K to 128K. (3) Latency: cut-through latency dropped from 5µs to 1.2µs at 10G. (4) Power efficiency: 0.8W per 10G port vs. 1.5W previously. For 40G/100G uplink models, backplane changed from distributed fabric to single-stage Clos, improving oversubscription ratios from 4:1 to 1:1 on line-rate ports.

Q4: What is the basic CLI configuration steps to enable a trunk port between two custom fiber optic switches?

Step 1: Enter config mode: 'configure terminal'. Step 2: Specify interface: 'interface ethernet 1/1/1'. Step 3: Set to trunk mode: 'switchport mode trunk'. Step 4: Allow VLANs: 'switchport trunk allowed vlan add 10,20,30'. Step 5: (Optional) Set native VLAN: 'switchport trunk native vlan 99'. Step 6: Enable the port: 'no shutdown'. Step 7: Verify: 'show interface trunk'. For LACP bundling, precede with 'interface port-channel 1' then 'channel-group 1 mode active' per physical interface. Always commit changes ('write memory') after testing.

Q5: How do you configure virtual stacking (cluster) across multiple custom fiber optic switches?

Our custom fiber optic switches support up to 12 units in a virtual chassis using dedicated stacking ports (QSFP or 40G ports). Configuration steps: (1) Assign a unique stack member ID (1-12) to each switch: 'stack member-id 2'. (2) Set priority (higher wins master election): 'stack priority 200'. (3) Designate stack ports: 'stack port 1 interface ethernet 1/1/49', 'stack port 2 interface ethernet 1/1/50'. (4) Connect stack cables in a ring topology (no single point of failure). (5) Enable stacking: 'stack enable'. (6) Verify with 'show stack topology' and 'show stack members'. Stacking bandwidth reaches 320 Gbps per switch, with sub-50ms failover if the master fails.

Q6: What troubleshooting steps for redundant power supply (RPS) failover on a custom fiber optic switch?

First, confirm both PSUs are inserted and powered: 'show power-supply' should list PSU1 as 'Active' and PSU2 as 'Standby' or 'Active'. If failover doesn't occur: (1) Verify PSU2 has input voltage (110/220V AC or -48V DC) using a multimeter. (2) Check PSU2 status LED: green means ready, amber indicates fault. (3) Examine logs: 'show logging' for 'PSU2 failed' or 'input under-voltage'. (4) Test forced failover: 'redundancy force-switchover' – the switch should remain up without packet loss. (5) If failover fails, update firmware (some early revisions had PSU hot-swap detection bugs). (6) For persistent issues, check backplane PSU connector for bent pins. Always keep one PSU as cold-spare on shelf.

Q7: What is the maximum throughput in Gbps for a 48x10G + 4x100G custom fiber optic switch under full line-rate load?

Maximum throughput is 1.28 Tbps (bidirectional). Calculation: 48 ports × 10 Gbps × 2 (full duplex) = 960 Gbps, plus 4 ports × 100 Gbps × 2 = 800 Gbps, total 1.76 Tbps raw. However, the ASIC fabric limits non-blocking to 1.28 Tbps, meaning the four 100G uplinks share 1.28 Tbps bandwidth. Under real-world 1518-byte packets, forwarding rate sustains 106 Mpps. For micro-burst absorption, the 32MB packet buffer handles 200ms of traffic at 1.28 Tbps before tail-drop occurs. To achieve full 1:1 oversubscription, specify a custom backplane upgrade to 1.76 Tbps (engineering lead time +14 days).