如何解决错误:主机to域(不存在域)的epmd错误？

Question

我正在尝试使用这个RabbitMQ在裸金属上在K8S集群上设置K8S运算符和链接集群。

K8S集群有一个主节点和一个工作节点

RabbitMQ集群吊舱日志

root@re-ctrl01 01 containers# kubectl日志定义-server-0 -n rabbitmq-system

BOOT FAILED  (Tailored output)
===========
ERROR: epmd error for host definition-server-0.definition-nodes.rabbitmq-system: nxdomain (non-existing domain)

11:51:13.733 [error] Supervisor rabbit_prelaunch_sup had child prelaunch started with rabbit_prelaunch:run_prelaunch_first_phase() at undefined exit with reason {epmd_error,"definition-server-0.definition-nodes.rabbitmq-system",nxdomain} in context start_error.  Crash dump is being written to: erl_crash.dump...

root@ctrl01 01 containers# kubectl描述荚定义-server-0 -n rabbitmq-system

Name:         definition-server-0
Namespace:    rabbitmq-system
Priority:     0
Events:
  Type     Reason     Age               From               Message
  ----     ------     ----              ----               -------
  Normal   Scheduled  44s               default-scheduler  Successfully assigned rabbitmq-system/definition-server-0 to re-ctrl01.local
  Normal   Pulled     43s               kubelet            Container image "rabbitmq:3.8.16-management" already present on machine
  Normal   Created    43s               kubelet            Created container setup-container
  Normal   Started    43s               kubelet            Started container setup-container
  Normal   Pulled     42s               kubelet            Container image "rabbitmq:3.8.16-management" already present on machine
  Normal   Created    42s               kubelet            Created container rabbitmq
  Normal   Started    42s               kubelet            Started container rabbitmq
  Warning  Unhealthy  4s (x3 over 24s)  kubelet            Readiness probe failed: dial tcp 10.244.0.xxx:5672: connect: connection refused

我将以下条目添加到worker node的worker node文件中，因为我不确定该条目是否必须添加到master or worker中

[root@re-worker01 ~]# cat /etc/hosts
127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4
127.0.0.1   re-worker01.local re-worker01 definition-server-0.definition-nodes.rabbitmq-system

我被这个问题困扰了将近两天。我搜索并发现了类似的问题，但没有解决我的问题。

我在pod日志中看到了多个问题，并描述了输出，我无法找出根本原因。

1. 在哪里可以在erl_crash.dump上找到K8S文件？
2. 这真的是与主机名有关的问题吗？
3. 10.244.0.xxx:5672: connection :connection拒绝--这个问题是因为'epmd'还是其他原因？

Answer 1

我花了很多时间才设法解决了这个问题。

我使用definition-server-0.definition-nodes.rabbitmq-system /etc/hosts hostAliases将主机添加到RabbitMQ集群荚的文件中。

下面给出了添加hostAliases的YAML

apiVersion: rabbitmq.com/v1beta1
kind: RabbitmqCluster
metadata:
  name: definition
  namespace: rabbitmq-system
spec:
  replicas: 1
  override:
    statefulSet:
      spec:
        template:
          spec:
            containers: []
            hostAliases:
            - ip: "127.0.0.1"
              hostnames:
              - "definition-server-0"
              - "definition-server-0.definition-nodes.rabbitmq-system"