难以理解在ESX VM上具有Kubernetes群集的节点上的dmesg输出

问题描述 投票:0回答:1

由于kubespray,我配置了在ESX上的VM上运行的Kubernetes测试群集。在配置文件中,我告诉Kubespray将带有calico的群集配置为具有默认CIDR的CNI。

[当我使用vSphere访问VM时,或者在我的VM上运行命令dmesg时,我的主数据库上都会显示此输出:

[2866556.027837] IPVS: rr: TCP 10.233.13.12:443 - no destination available
[2866556.857464] IPVS: rr: TCP 10.233.13.12:443 - no destination available
[2866557.029471] IPVS: rr: TCP 10.233.13.12:443 - no destination available
[2866688.881160] IPVS: __ip_vs_del_service: enter
[2866689.018851] IPVS: __ip_vs_del_service: enter
[2866689.023030] IPVS: __ip_vs_del_service: enter
[2866689.188072] IPVS: __ip_vs_del_service: enter
[2866689.416153] IPVS: __ip_vs_del_service: enter
[2866689.420334] IPVS: __ip_vs_del_service: enter
[2866692.005599] IPVS: __ip_vs_del_service: enter
[2866692.010260] IPVS: __ip_vs_del_service: enter
[2866692.257045] IPVS: __ip_vs_del_service: enter
[2866692.265034] IPVS: __ip_vs_del_service: enter
[2866692.267455] IPVS: __ip_vs_del_service: enter
[2866692.267493] IPVS: __ip_vs_del_service: enter
[2866916.815472] IPVS: rr: TCP 10.233.49.127:443 - no destination available
[2866916.820841] IPVS: rr: TCP 10.233.49.127:443 - no destination available
[2866916.823418] IPVS: rr: TCP 10.233.49.127:443 - no destination available
[2866916.824167] IPVS: rr: TCP 10.233.49.127:443 - no destination available
[2866916.826243] IPVS: rr: TCP 10.233.49.127:443 - no destination available

以及我的工作者的输出

[1207664.350374] IPVS: rr: TCP 10.233.3.61:8080 - no destination available
[1207664.422584] IPVS: rr: TCP 10.233.3.61:8080 - no destination available
[1207667.108560] net_ratelimit: 13 callbacks suppressed
[1207667.108567] IPVS: rr: TCP 10.233.3.61:8080 - no destination available
[1207667.217235] IPVS: rr: TCP 10.233.3.61:8080 - no destination available
[1207667.274593] IPVS: rr: TCP 10.233.3.61:8080 - no destination available
[1207667.331658] IPVS: rr: TCP 10.233.3.61:8080 - no destination available
[1207668.218597] IPVS: rr: TCP 10.233.3.61:8080 - no destination available
[1207668.334613] IPVS: rr: TCP 10.233.3.61:8080 - no destination available
[1207675.500914] IPVS: rr: TCP 10.233.49.141:8086 - no destination available
[1207676.502566] IPVS: rr: TCP 10.233.49.141:8086 - no destination available
[1207676.628377] IPVS: rr: TCP 10.233.49.141:8086 - no destination available
[1208009.456587] blk_update_request: I/O error, dev fd0, sector 0
[1208009.924355] blk_update_request: I/O error, dev fd0, sector 0
[1208058.699578] blk_update_request: I/O error, dev fd0, sector 0
[1208240.706522] IPVS: Creating netns size=2048 id=289
[1208241.432437] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[1208241.445496] IPv6: ADDRCONF(NETDEV_UP): cali6ef7aa1f11f: link is not ready
[1208241.447406] IPv6: ADDRCONF(NETDEV_CHANGE): cali6ef7aa1f11f: link becomes ready
[1208241.447469] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready

我真的很难理解那些日志。它似乎与带有IPVS : rr消息的calico链接,并且每个IP地址都对应于群集中的服务IP。但是我没有在清单文件中的[calico-rr]下配置任何节点,因为它是可选的,它可用于改善大型集群之间的BGP通信。

[all]
m1 ansible_host=x.x.x.x ip=x.x.x.x
m2 ansible_host=x.x.x.x ip=x.x.x.x
w1 ansible_host=x.x.x.x ip=x.x.x.x
w2 ansible_host=x.x.x.x ip=x.x.x.x
w3 ansible_host=x.x.x.x ip=x.x.x.x

[kube-master]
m1
m2

[etcd]
m1
m2
w1

[kube-node]
w1
w2
w3

[calico-rr]

[k8s-cluster:children]
kube-master
kube-node
calico-rr

据我所知,当我应用yaml文件安装linkerd时,这是在配置新Pod和服务期间出现的输出。这与准备情况调查有联系吗?消息弹出,直到服务/窗格准备就绪?

真正的问题是这些日志在vSphere的控制台上发送了垃圾邮件,我真的不知道如何清除它们。

我在其他线程上搜索了更多信息,但发现并没有太大帮助。

更新:

我对IPVS有更多见解:rr错误。https://kubernetes.io/blog/2018/07/09/ipvs-based-in-cluster-load-balancing-deep-dive/它链接到kube-proxy使用的IPVS负载平衡。

但是我仍然不知道如何通过vSphere在我的ESX控制台上没有这些日志。

更新2:

对于使用Kubespray进行Kubernetes安装,我只是按照下面的指南进行操作,并如上所述更改了清单文件。

VM OS:Centos 7.7 1908

Kubernetes版本:1.16.3

Kubespray版本:release-2.12

Kubespray入门指南:https://github.com/kubernetes-sigs/kubespray/blob/master/docs/getting-started.md

kubernetes vmware vsphere calico kubespray
1个回答
0
投票
© www.soinside.com 2019 - 2024. All rights reserved.