documentation:examples:ipsec_performance_of_a_netgate_rcc-ve_4860
Differences
This shows you the differences between two versions of the page.
| Both sides previous revisionPrevious revisionNext revision | Previous revision | ||
| documentation:examples:ipsec_performance_of_a_netgate_rcc-ve_4860 [2020/09/22 12:02] – [Netgate (DUT)] olivier | documentation:examples:ipsec_performance_of_a_netgate_rcc-ve_4860 [2020/09/22 12:08] (current) – [Using IPSec bench "Equilibrium throughput" method] olivier | ||
|---|---|---|---|
| Line 100: | Line 100: | ||
| ==== R3 (Reference device) ==== | ==== R3 (Reference device) ==== | ||
| + | |||
| + | / | ||
| + | < | ||
| + | # Loading AES-NI module sooner to be sure it is loaded before IPsec keys | ||
| + | aesni_load=" | ||
| + | </ | ||
| Configure IP address, routes and static IPSec. | Configure IP address, routes and static IPSec. | ||
| Line 107: | Line 113: | ||
| # IPv4 router | # IPv4 router | ||
| gateway_enable=" | gateway_enable=" | ||
| - | ifconfig_igb2=" | + | ifconfig_igb2=" |
| - | ifconfig_igb3=" | + | ifconfig_igb3=" |
| static_routes=" | static_routes=" | ||
| - | route_generator=" | + | route_generator=" |
| route_receiver=" | route_receiver=" | ||
| static_arp_pairs=" | static_arp_pairs=" | ||
| static_arp_generator=" | static_arp_generator=" | ||
| - | static_arp_receiver=" | + | static_arp_receiver=" |
| # IPv6 router | # IPv6 router | ||
| Line 124: | Line 130: | ||
| ipv6_static_routes=" | ipv6_static_routes=" | ||
| - | ipv6_route_generator=" | + | ipv6_route_generator=" |
| ipv6_route_receiver=" | ipv6_route_receiver=" | ||
| static_ndp_pairs=" | static_ndp_pairs=" | ||
| static_ndp_generator=" | static_ndp_generator=" | ||
| - | static_ndp_receiver=" | + | static_ndp_receiver=" |
| + | cloned_interfaces=" | ||
| + | create_args_ipsec0=" | ||
| + | ifconfig_ipsec0=" | ||
| + | ifconfig_ipsec0_ipv6=" | ||
| - | # Enabling | + | # Enabling |
| - | kld_list=" | + | |
| ipsec_enable=" | ipsec_enable=" | ||
| </ | </ | ||
| Line 140: | Line 149: | ||
| flush; | flush; | ||
| spdflush; | spdflush; | ||
| - | spdadd 198.18.0.0/ | + | add 198.18.1.203 198.18.1.209 esp 10000 -m tunnel -u 200 -E aes-gcm-16 " |
| - | spdadd 198.19.0.0/ | + | add 198.18.1.209 198.18.1.203 esp 10001 -m tunnel -u 200 -E aes-gcm-16 " |
| - | add 198.18.1.203 198.18.1.209 esp 0x1000 | + | |
| - | add 198.18.1.209 198.18.1.203 esp 0x1001 | + | |
| - | spdadd 2001:2::/49 2001: | + | |
| - | spdadd 2001: | + | |
| - | add 2001: | + | |
| - | add 2001: | + | |
| </ | </ | ||
| ===== Using IPSec bench " | ===== Using IPSec bench " | ||
| Once done, we start using a fast method for measuring the "IPsec equilibrium throughput" | Once done, we start using a fast method for measuring the "IPsec equilibrium throughput" | ||
| - | |||
| - | Notice that the reference device (IBM x3550-M3) used in front of the Netgate has a [[IPSec performance lab of an IBM System x3550 M3 with Intel 82580|equilibrium throughput of 843Mb/s]]. Then if the value measured during this bench is close to 843Mb/s we had to found a more powerful reference device. | ||
| From the packet generator/ | From the packet generator/ | ||
| < | < | ||
| - | [root@R1]~# equilibrium | + | [root@R1]~# equilibrium -4 -u -d 00: |
| Benchmark tool using equilibrium throughput method | Benchmark tool using equilibrium throughput method | ||
| - Benchmark mode: Bandwitdh (bps) for VPN gateway | - Benchmark mode: Bandwitdh (bps) for VPN gateway | ||
| Line 166: | Line 167: | ||
| - Offering load = 500 Mb/s | - Offering load = 500 Mb/s | ||
| - Step = 250 Mb/s | - Step = 250 Mb/s | ||
| - | - Measured forwarding rate = 399 Mb/s | + | - Measured forwarding rate = 499 Mb/s |
| Iteration 2 | Iteration 2 | ||
| - | - Offering load = 250 Mb/s | + | - Offering load = 750 Mb/s |
| - Step = 250 Mb/s | - Step = 250 Mb/s | ||
| - | - Trend = decreasing | + | - Trend = increasing |
| - | - Measured forwarding rate = 250 Mb/s | + | - Measured forwarding rate = 670 Mb/s |
| Iteration 3 | Iteration 3 | ||
| - | - Offering load = 375 Mb/s | + | - Offering load = 625 Mb/s |
| - Step = 125 Mb/s | - Step = 125 Mb/s | ||
| - | - Trend = increasing | + | - Trend = decreasing |
| - | - Measured forwarding rate = 375 Mb/s | + | - Measured forwarding rate = 624 Mb/s |
| Iteration 4 | Iteration 4 | ||
| - | - Offering load = 437 Mb/s | + | - Offering load = 687 Mb/s |
| - Step = 62 Mb/s | - Step = 62 Mb/s | ||
| - Trend = increasing | - Trend = increasing | ||
| - | - Measured forwarding rate = 399 Mb/s | + | - Measured forwarding rate = 672 Mb/s |
| Iteration 5 | Iteration 5 | ||
| - | - Offering load = 406 Mb/s | + | - Offering load = 656 Mb/s |
| - Step = 31 Mb/s | - Step = 31 Mb/s | ||
| - Trend = decreasing | - Trend = decreasing | ||
| - | - Measured forwarding rate = 399 Mb/s | + | - Measured forwarding rate = 655 Mb/s |
| Iteration 6 | Iteration 6 | ||
| - | - Offering load = 391 Mb/s | + | - Offering load = 671 Mb/s |
| - Step = 15 Mb/s | - Step = 15 Mb/s | ||
| - | - Trend = decreasing | + | - Trend = increasing |
| - | - Measured forwarding rate = 391 Mb/s | + | - Measured forwarding rate = 670 Mb/s |
| Iteration 7 | Iteration 7 | ||
| - | - Offering load = 398 Mb/s | + | - Offering load = 678 Mb/s |
| - Step = 7 Mb/s | - Step = 7 Mb/s | ||
| - Trend = increasing | - Trend = increasing | ||
| - | - Measured forwarding rate = 398 Mb/s | + | - Measured forwarding rate = 670 Mb/s |
| - | Estimated Equilibrium Ethernet throughput= | + | Estimated Equilibrium Ethernet throughput= |
| </ | </ | ||
| - | Using AES-GCM-128 and aesni kernel module loaded on the NetGate RCC-VE 4860, we can estimate an IPSec Equilibrium throughput of about 400Mb/s. | + | Using AES-GCM-128 and aesni kernel module loaded on the NetGate RCC-VE 4860, we can estimate an IPSec Equilibrium throughput of about 672Mb/s. |
| - | {{bench-ipsec-netgate-12head.png}} | + | {{bench.netgate.ipsec.13head.png}} |
documentation/examples/ipsec_performance_of_a_netgate_rcc-ve_4860.1600768932.txt.gz · Last modified: 2020/09/22 12:02 by olivier