NetworkPolicy意外阻止Ingress流量问题求助

我碰到了一个NetworkPolicy导致的流量阻塞问题，折腾了半天没找到原因，来请教各位大佬：

先贴一下我的NetworkPolicy配置：

spec:
  podSelector:
    matchExpressions:
      - key: app.kubernetes.io/component
        operator: In
        values:
          - "backend"
          - "read"
          - "write"
          - "memcached-chunk"
          - "memcached-results"
  policyTypes:
    - Ingress
    - Egress
  ingress:
    - from:
        - namespaceSelector:
            matchExpressions:
              - key: kubernetes.io/metadata.name
                operator: In
                values:
                  - "{{ .Release.Namespace }}"
                  - "ingress-nginx"
        - ipBlock:
            cidr: 192.168.0.0/16
  egress:
    - {}

问题是这样的：我从192.168.126.128这个IP通过Ingress访问后端API时，请求直接超时了。查看nginx ingress-controller的日志，发现了对应的错误：

2024/05/14 17:23:27 [error] 5437#5437: *6517485 upstream timed out (110: Operation timed out) while connecting to upstream...

我本来以为Ingress所在的ingress-nginx命名空间已经被允许访问后端Pod，而且客户端IP也在192.168.0.0/16网段里，但为什么流量还是被阻断了呢？麻烦帮我分析一下原因，谢谢！

备注：内容来源于stack exchange，提问作者joker57