Within the same private network, there are three Linux servers: 172.168.0.1, 172.168.0.2, and 172.168.0.3. 172.168.0.1 is the pressure testing machine running Jemeter, 172.168.0.2 is the traifik node, and 172.168.0.3 is the server running pod. When starting the pressure test, using tcpdump to capture packets, 172.168.0.2 can capture packets, but 172.168.0.3 did not receive packets, The Jemeter error message is as follows: 500 Internal Server Error and 502 bad gateway. But if nginx is deployed and proxied to the nodeport of k8s through nginx, the pressure test is normal.