cloudpods
cloudpods copied to clipboard
yunionio
[BUG]把一个EIP解邦后再邦定给另外一个VPC 的虚拟机， 会出现ping 这个EIP 时断的情况 。
问题描述/What happened: 把一个EIP解邦后再邦定给另外一个VPC 的虚拟机， 会出现ping 这个EIP 时断的情况 ，通过抓包分析发现解绑或者删除EIP 存在清理不全的情况 环境/Environment: EIP 网关是独立部署的， 如果EIP 不存在重复使用的情况下， 网络是正常的。
- OS (e.g.
cat /etc/os-release):
[root@ecm-232e-0001 ~]# cat /etc/os-release NAME="CentOS Linux" VERSION="7 (Core)" ID="centos" ID_LIKE="rhel fedora" VERSION_ID="7" PRETTY_NAME="CentOS Linux 7 (Core)" ANSI_COLOR="0;31" CPE_NAME="cpe:/o:centos:centos:7" HOME_URL="https://www.centos.org/" BUG_REPORT_URL="https://bugs.centos.org/"
CENTOS_MANTISBT_PROJECT="CentOS-7" CENTOS_MANTISBT_PROJECT_VERSION="7" REDHAT_SUPPORT_PRODUCT="centos" REDHAT_SUPPORT_PRODUCT_VERSION="7"
[root@ecm-232e-0001 ~]#
- Kernel (e.g.
uname -a):
[root@ecm-232e-0001 ~]# uname -a Linux ecm-232e-0001 5.4.130-1.yn20230805.el7.x86_64 #1 SMP Wed Oct 11 03:26:01 UTC 2023 x86_64 x86_64 x86_64 GNU/Linux [root@ecm-232e-0001 ~]# [root@ecm-232e-0001 ~]#
- Host: (e.g.
dmidecode | egrep -i 'manufacturer|product' |sort -u)
[root@ecm-232e-0001 ~]# dmidecode | egrep -i 'manufacturer|product' |sort -u Manufacturer: GoStack Foundation Manufacturer: Red Hat Product Name: Gostack Agent [root@ecm-232e-0001 ~]#
- Service Version (e.g.
kubectl exec -n onecloud $(kubectl get pods -n onecloud | grep climc | awk '{print $1}') -- climc version-list):
[root@ecm-232e-0001 ~]# climc version-list Get "https://192.168.1.3:30898/version": dial tcp 192.168.1.3:30898: connect: connection refused Get "https://192.168.1.3:30893/version": dial tcp 192.168.1.3:30893: connect: connection refused Get "https://192.168.1.3:30892/version": dial tcp 192.168.1.3:30892: connect: connection refused Get "https://192.168.1.3:30443/version": dial tcp 192.168.1.3:30443: connect: connection refused +---------------+--------------------------------------------+ | Field | Value | +---------------+--------------------------------------------+ | ansible | release/3.11(3ffea07d3124042402) | | apimap | release/3.11(3ffea07d3124042402) | | cloudmon | release/3.11(3ffea07d3124042402) | | cloudproxy | release/3.11(3ffea07d3124042402) | | compute_v2 | release/3.11(3ffea07d3124042402) | | devtool | release/3.11(3ffea07d3124042402) | | identity | release/3.11(3ffea07d3124042402) | | image | release/3.11(3ffea07d3124042402) | | k8s | heads/v3.11.3-20240422.2(e6c3e48724042402) | | log | release/3.11(3ffea07d3124042402) | | monitor | release/3.11(3ffea07d3124042402) | | notify | release/3.11(3ffea07d3124042402) | | scheduledtask | release/3.11(3ffea07d3124042402) | | scheduler | release/3.11(3ffea07d3124042402) | | vpcagent | release/3.11(3ffea07d3124042402) | | webconsole | release/3.11(3ffea07d3124042402) | | yunionconf | release/3.11(3ffea07d3124042402) | +---------------+--------------------------------------------+ [root@ecm-232e-0001 ~]#
以 21.zip 下是通过抓包分析的情况
这是从 EIP 网关抓的包
只要重新启动 EIP 网关 yunion-sdnagent-eipgw ， 就正常。
我把这个问题复现了
删除 虚拟机的时候， 如不不选择删除EIP ， 后面再创建虚拟机， 选择之前的EIP，就有问题了
这是 在 EIP 网关查询出来的 flows 记录 ， 对应的vpc 地址是错误，
重新启动 systemctl restart yunion-sdnagent-eipgw ，flows 记录 就对了