Optimizing NP6 performance by distributing traffic to XAUI links
On FortiGate units with NP6 processors, the FortiGate interfaces are switch ports that connect to the NP6 processors with XAUI links. Each NP6 processor has a 40-Gigabit bandwidth capacity. Traffic passes from the interfaces to each NP6 processor over four XAUI links. The four XAUI links each have a 10-Gigabit capacity for a total of 40 Gigabits.
On many FortiGate units with NP6 processors, the NP6 processors and the XAUI links are over-subscribed. Since the NP6 processors are connected by an Integrated Switch Fabric, you do not have control over how traffic is distributed to them. In fact traffic is distributed evenly by the ISF.
However, you can control how traffic is distributed to the XAUI links and you can optimize performance by distributing traffic evenly among the XAUI links. For example, if you have a very high amount of traffic passing between two networks, you can connect each network to interfaces connected to different XAUI links to distribute the traffic for each network to a different XAUI link.
Example FortiGate 3300E
On the FortiGate 3300E (See FortiGate 3300E and 3301E fast path architecture), there are 34 data interfaces of various speeds that send and receive traffic for four NP6 processors over a total of sixteen 10-Gigabit XAUI links. The amount of traffic that the FortiGate 3300E can offload is limited by the number of NP6 processors and the number of XAUI links. You can optimize the amount of traffic that the FortiGate 3300E can process by distributing it evenly amount the XAUI links and the NP6 processors.
You can see the FortiGate 3300E Ethernet interface, XAUI link, and NP6 configuration by entering the get hardware npu np6 port-list command. For the FortiGate 3300E the output is:
get hardware npu np6 port-list
Chip XAUI Ports Max Cross-chip
Speed offloading
------ ---- ------- ----- ----------
np6_0 0 port1 1G Yes
0 port14 10G Yes
1 port2 1G Yes
1 port15 10G Yes
2 port3 1G Yes
2 port16 10G Yes
3 port13 10G Yes
0-3 port17 25G Yes
0-3 port31 40G Yes
------ ---- ------- ----- ----------
np6_1 0 port4 1G Yes
1 port5 1G Yes
2 port6 1G Yes
3
0-3 port18 25G Yes
0-3 port19 25G Yes
0-3 port20 25G Yes
0-3 port24 25G Yes
0-3 port23 25G Yes
0-3 port32 40G Yes
------ ---- ------- ----- ----------
np6_2 0 port7 1G Yes
1 port8 1G Yes
2 port9 1G Yes
3
0-3 port22 25G Yes
0-3 port21 25G Yes
0-3 port26 25G Yes
0-3 port25 25G Yes
0-3 port28 25G Yes
0-3 port33 40G Yes
------ ---- ------- ----- ----------
np6_3 0 port10 1G Yes
1 port11 1G Yes
2 port12 1G Yes
2 port29 10G Yes
3 port30 10G Yes
0-3 port27 25G Yes
0-3 port34 40G Yes
------ ---- ------- ----- ----------
In this command output you can see that each NP6 has four XAUI links (0 to 3) and the mapping between XAUI ports and interfaces is different for each NP6 processor.
NP6_0 has the following XAUI mapping:
- port1 (1G) and port14 (10G) are connected to XAUI link 0.
- port2 (1G) and port15 (10G) are connected to XAUI link 1.
- port3 (1G) and port16 (10G) are connected to XAUI link 2.
- port13 (10G) is connected to XAUI link 3.
- port17 (25G) and port31 (40G) are connect to all four of the XAUI links (0-3).
The interfaces connected to NP6_0 have a total capacity of 108G, but NP6_0 has total capacity of 40G. For optimal performance, no more than 40G of this capacity should be used or performance will be affected. For example, if you connect port31 to a busy 40G network you should avoid using any of the other ports connected to NP6_0. If you connect port17 to a 25G network, you can also connect one or two 10G interfaces (for example, port14 and 15). You can connect port13, port14, port15, and port16 to four 10G networks if you avoid using any of the other interfaces connected to NP6_0.