NSH_SFC 17.01 Performance Report Summary

The NSH_SFC 17.01 Performance Report focuses on measuring and analyzing the performance of various elements such as Service Function Forwarder, NSH Proxy, NSH Classifier, and more in the context of VPP 17.01 for different SFC ingredients. Baseline performance is established using IXIA-based PacketGen, with future testing planned using TREX automation. The report also discusses testing configurations, device under test details, and performance benchmarks for NSH_SFC throughput. Overall, the document provides an in-depth look at the performance metrics and testing methodologies used by the CSIT team for NSH_SFC 17.01.

• NSH SFC
• Performance Report
• VPP 17.01
• CSIT Team
• Automation Testing

heatherly_e Follow

Uploaded on Sep 28, 2024 | 0 Views

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

Presentation Transcript

1. NSH_SFC 17.01 Performance Report FD.io NSH_SFC and CSIT Team INTEL CONFIDENTIAL

2. Background Background Measure NSH_SFC 17.01 performance with VPP 17.01 for different SFC ingredients Service Function Forwarder NSH Proxy NSH Classifier IXIA based PacketGen to be replaced by TREX in the automation test by CSIT team Baseline performance will be used to analyze potential negative performance impact introduced by software PacketGen or future patches As baseline for apple-to-apple performance comparison with DPDK OVS, once NSH patches to OVS are upstreamed Focus on single core performance currently, scale-up (assign more cores) testing will be covered in the future

3. Test Configuration Test Configuration IXIA 10G Niantic Core 2 Core 3 Core 0 Core 1 Port 1 2.2GHz Port B DIMM 0 DIMM 1 Port 0 Port A DIMM 2 DIMM 3 IXIA Traffic Configuration NSH Classifier: Raw packet in different size Inner MAC header inner IP header L4 packet Inner MAC header Outer IP header Outer MAC header inner IP header UDP header SFF: VxLAN-GPE + NSH + Raw Packet VxLAN-GPE Header L4 packet NSH Outer MAC header inner IP header Inner MAC header Outer IP header NSH Proxy Inbound: VxLAN-GPE + NSH + Raw Packet UDP header VxLAN-GPE Header L4 packet NSH Outer IP header Outer MAC header inner IP header Inner MAC header UDP header NSH Proxy Outbound: VxLAN+ Raw Packet VxLAN L4 packet

4. Device Under Test Device Under Test BIOS Configuration Hardware Configuration Intel(R) Xeon(R) CPU E5-2699 v4 @ 2.20GHz Enhanced Intel Speedstep Disabled CPU Turbo Boost Disabled DIMM 2133 MHz, 64GB Total Processor C3 Disabled Processor C6 Disabled 2x 82599ES 10-Gigabit SFI/SFP+ Network Connection NIC Hyper-Threading Disabled Intel VT-d Enabled Ixia* 10 Gigabit Ethernet Traffic Generator (16 ports) CPU Power and Performance Policy Performance PacketGen Memory Freq. 2133 MHz Total Memory Size 64 GB Software Configuration Memory RAS and Performance Configuration -> NUMA Optimized ENABLED Ubuntu 14.04.1 QPI B/W 9.6 GT/s OS MLC Streamer ENABLED 4.2.0-27-generic Kernel version MLC Spatial Prefetcher ENABLED vpp v1.0.0-263~gd6f5b96 VPP DCU Data Prefetcher ENABLED DPDK 16.11 DPDK DCU Instruction Prefetcher ENABLED 17.01 RC2 NSH_SFC Direct Cache Access (DCA) ENABLED 4

5. NSH_SFC Performance in 17.01 Release NSH_SFC Throughput For 128B Packet (1C1T) NSH Classifier Throughput For Different Packet Size (1C1T) 10000 8000 9432 7088 9000 7000 8000 Throughput (Mbps) 7088 6000 Throughput (Mbps) 7000 5000 6000 5520 4500 4250 5120 5000 4000 4000 3000 3000 2000 2000 1000 1000 0 0 72 84 128 256 Packet Size (Bytes) NSH Classifier NSH Proxy (inbound) SFF NSH_SFC 17.01 performance drops from 5.8Mpps to 4.25Mpps for NSH Proxy 2 to 3 dedicated cores could achieve 10G line rate for small packet for all NSH_SFC use cases

6. Backup Backup

7. NSH Proxy Performance in 16.09 Release - As Reference Data NSH-VxLAN-GPE Decap 7 VxLAN Encap nsh-proxy uni-direction performance(nsh-decap->vxlan-encap vs. vxlan-decap->nsh-encap) VPP 6.01 6.01 VPP NSH-Proxy 5.83 5.83 6 DPDK Port Throughput (Mpps) 5 Nsh-vxlan-gpe Pkts VxLAN Pkts 3.83 3.83 4 IXIA port 3 Figure 3. nsh-proxy-decap/Uni 2.15 2.15 dpdk/ethernet input 2 1.49 1.49 ip4-input 1.14 1.14 0.93 0.93 0.81 0.81 1 ip4-local 0 Ip4-udp-lookup 64 80 114 128 178 306 562 818 1074 1330 1518 vxlan-gpe-nsh-decap->vxlan-encap 5.83 5.83 3.83 2.15 1.49 1.14 0.93 0.81 nsh-vxlan-gpe-input vxlan-decap->vxlan-gpe-nsh-encap 6.01 6.01 3.83 2.15 1.49 1.14 0.93 0.81 nsh-proxy-input vxlan-encap nsh-proxy-decap for 128B Ip4-lookup CPP 394.9 Time (ns) 171.7 Mpps 5.83 Ip4-rewrite-transmit dpdk port-output dpdk port-tx