Calico单一pod not ready
问题描述：
自建K8S，查看Calico的pod状态，发现node1上面的calico没有ready
[root@master ec2-user]# kubectl -n kube-system get pod -o wide| grep cali
calico-kube-controllers-677cd97c8d-2rxtd 1/1 Running 0 18h 10.244.135.1 node3
calico-node-jw575 1/1 Running 0 18h 10.0.2.50 node4
calico-node-mkcwg 1/1 Running 0 18h 172.31.24.53 node3
calico-node-v8zph 1/1 Running 0 18h 10.0.0.180 master
calico-node-xrq8m 1/1 Running 0 18h 10.0.0.106 node2
calico-node-xtg8k 0/1 Running 0 16h 10.0.0.27 node1 分析过程：
按理说都是一个yaml部署的，不应该有问题，重建了一下pod还是这个情况，看一下事件。
Normal Created 5m19s kubelet Created container calico-node
Warning Unhealthy 5m17s kubelet Readiness probe failed: calico/node is not ready: felix is not ready: readiness probe reporting 503
Warning Unhealthy 5m16s (x2 over 5m18s) kubelet Readiness probe failed: calico/node is not ready: BIRD is not ready: Error querying BIRD: unable to connect to BIRDv4 socket: dial unix /var/run/calico/bird.ctl: connect: connection refused
Warning Unhealthy 5m11s kubelet Readiness probe failed: 2025-07-22 11:56:14.088 [INFO][210] confd/health.go 180: Number of node(s) with BGP peering established = 0
calico/node is not ready: BIRD is not ready: BGP not established with 10.0.0.180,10.0.0.106,172.31.24.53,10.0.2.50
Warning Unhealthy 5m1s kubelet Readiness probe failed: 2025-07-22 11:56:24.104 [INFO][230] confd/health.go 180: Number of node(s) with BGP peering established = 0
calico/node is not ready: BIRD is not ready: BGP not established with 10.0.0.180,10.0.0.106,172.31.24.53,10.0.2.50看来主要的是BGP邻居没有建立成功。看一下calico的日志，没发现什么问题。bird是用于BGP的进程。
bird: Mesh_10_0_0_180: Starting
bird: Mesh_10_0_0_180: State changed to start
bird: Mesh_10_0_0_106: Starting
bird: Mesh_10_0_0_106: State changed to start
bird: Mesh_172_31_24_53: Starting
bird: Mesh_172_31_24_53: State changed to start
bird: Mesh_10_0_2_50: Starting
bird: Mesh_10_0_2_50: State changed to start在节点上抓了一下179端口的报文，发现三次握手能建立，但是发送OPEN报文之后就被RST了。值得注意的是，是node1的calico发送报文给其他node，被其他node给RST了，怀疑BGP的参数有问题，这下只能登陆进去看看了。
登录进node1和另一个node上面的calico看一下BGP配置
node1的calico
[root@node1 /]# birdcl
BIRD v0.3.3+birdv1.6.8 ready.
bird> show protocols
****
Mesh_10_0_0_180 BGP master down 13:15:09 active
Mesh_10_0_0_106 BGP master down 13:15:09 active
Mesh_172_31_24_53 BGP master down 13:15:07 active
Mesh_10_0_2_50 BGP master down 13:15:08 active 另一个node的calico
Mesh_10_0_0_180 BGP master up 11:56:12 Established
Mesh_172_31_24_53 BGP master up 11:56:13 Established
Mesh_172_21_0_1 BGP master down 11:56:14 active
Mesh_10_0_0_27 BGP master up 13:15:09 Established 发现了问题，其他node要跟一个172.21.0.1的IP地址建立BGP，其他的IP地址都是node的IP，但node1的IP并不是这个172.21.0.1。下一步要看看这个IP怎么出来的。
还是要看node1上面calico的日志，发现了是从一个网卡上自动获取的。
[INFO][8] startup/startup.go 768: Using autodetected IPv4 address on interface br-18f903dae19f: 172.21.0.1/16看了下node1上确实有这个网卡
[root@node1 ec2-user]# ip add show
1: lo: mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
2: eth0: mtu 9001 qdisc mq state UP group default qlen 1000
link/ether 0a:7d:ab:6f:86:6a brd ff:ff:ff:ff:ff:ff
inet 10.0.0.27/24 brd 10.0.0.255 scope global dynamic eth0
valid_lft 2040sec preferred_lft 2040sec
3: docker0: mtu 1500 qdisc noqueue state DOWN group default
link/ether 02:42:9e:e6:d3:9d brd ff:ff:ff:ff:ff:ff
inet 172.17.0.1/16 brd 172.17.255.255 scope global docker0
valid_lft forever preferred_lft forever
4: tunl0@NONE: mtu 8981 qdisc noqueue state UNKNOWN group default qlen 1000
link/ipip 0.0.0.0 brd 0.0.0.0
inet 10.244.166.128/32 scope global tunl0
valid_lft forever preferred_lft forever
388: br-cf273aa4db36: mtu 1500 qdisc noop state DOWN group default
link/ether 02:42:68:c5:20:f6 brd ff:ff:ff:ff:ff:ff
inet 172.18.0.1/16 brd 172.18.255.255 scope global br-cf273aa4db36
valid_lft forever preferred_lft forever
389: br-3e57ff82d79a: mtu 1500 qdisc noop state DOWN group default
link/ether 02:42:e3:34:ee:59 brd ff:ff:ff:ff:ff:ff
inet 172.19.0.1/16 brd 172.19.255.255 scope global br-3e57ff82d79a
valid_lft forever preferred_lft forever
390: br-fbeba96c8bcb: mtu 1500 qdisc noop state DOWN group default
link/ether 02:42:86:31:fe:73 brd ff:ff:ff:ff:ff:ff
inet 172.20.0.1/16 brd 172.20.255.255 scope global br-fbeba96c8bcb
valid_lft forever preferred_lft forever
391: br-18f903dae19f: mtu 1500 qdisc noop state DOWN group default
link/ether 02:42:3a:7e:04:1d brd ff:ff:ff:ff:ff:ff
inet 172.21.0.1/16 brd 172.21.255.255 scope global br-18f903dae19f
valid_lft forever preferred_lft forever解决方案
回顾了一下calico的配置文件，没有单独指定过网卡； 对比了其他node，其他node确实没有这些br-的网卡，这个网卡应该是之前卸载flannel的时候没有卸载干净。我把这些网卡删除之后，calico就可以获取到eth0的IP地址，BGP就可以正常建立了。pod就ready了。
不过我还是想知道为什么自动获取没有获取到eth0的网卡，翻了一下calico的代码。
函数FilteredEnumeration从网卡列表中选择第一个可用的网卡以及它的IP
https://github.com/projectcalico/calico/blob/master/node/pkg/lifecycle/startup/autodetection/filtered.go#L30
// Find the first interface with a valid matching IP address and network.
// We initialise the IP with the first valid IP that we find just in
// case we don't find an IP *and* network.
for _, i := range interfaces {
log.WithField("Name", i.Name).Debug("Check interface")
for _, c := range i.Cidrs {
log.WithField("CIDR", c).Debug("Check address")
if c.IP.IsGlobalUnicast() && matchCIDRs(c.IP, cidrs) {
return &i, &c, nil
}
}
}那么这个网卡列表是怎么来的？有一个GetInterfaces函数，这个函数获取所有网卡列表
https://github.com/projectcalico/calico/blob/master/node/pkg/lifecycle/startup/autodetection/interfaces.go
// Loop through interfaces filtering on the regexes. Loop in reverse
// order to maintain behavior with older versions.
for idx := len(netIfaces) - 1; idx >= 0; idx-- {
iface := netIfaces[idx]
include := (includeRegexp == nil) || includeRegexp.MatchString(iface.Name)
exclude := (excludeRegexp != nil) && excludeRegexp.MatchString(iface.Name)
if include && !exclude {
if i, err := convertInterface(&iface, version); err == nil {
filteredIfaces = append(filteredIfaces, *i)
}
}
}这下找到原因了，这个地方的网卡队列顺序是倒着的，所以才没有首选eth0。
如果需要自己指定网卡，则可以在calico配置文件里添加如下环境变量。
# IP automatic detection
- name: IP_AUTODETECTION_METHOD
value: "interface=eth0"