no dns entry for VM - Githubissues

th-2021 commented 3 years ago

Is this a BUG REPORT or FEATURE REQUEST?:

Uncomment only one, leave it on its own line:

/kind bug /kind enhancement

What happened: New VMs with Debian 10 (buster)/Ubuntu 20.04 don't get a DNS entry (even if hostname and subdomain is set) What you expected to happen: As with other VMs (SLES/RHEL/Windows) a dns entry How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?:

Environment:

KubeVirt version (use virtctl version): 0.46.1
Kubernetes version (use kubectl version): 1.22
VM or VMI specifications:
Cloud provider or hardware configuration:
OS (e.g. from /etc/os-release): Ubuntu 20.04 (host)
Kernel (e.g. uname -a):
Install tools:
Others:

rhrazdil commented 3 years ago

Hello @th-2021 Could you provide us with the VM/VMI manifest used to create the VM?

th-2021 commented 3 years ago

apiVersion: kubevirt.io/v1alpha3
kind: VirtualMachine
metadata:
  labels:
    kubevirt.io/vm: ubuntu-home
  name: ubuntu-home
  namespace: vm-images
spec:
  runStrategy: "RerunOnFailure"
  template:
    metadata:
      labels:
        kubevirt.io/vm: ubuntu-home
      annotations:
        backup.velero.io/backup-volumes: datavolumedisk1
    spec:
      hostname: "ubuntu-home"
      subdomain: "vm"
      domain:
        devices:
          disks:
          - disk:
              bus: virtio
            name: datavolumedisk1
          - disk:
              bus: virtio
            name: cloudinitdisk
          interfaces:
          - bridge: {}
            name: default
          - bridge: {}
            name: ovs-blue-net
        machine:
          type: ""
        firmware:
          # this sets the bootloader type
          #bootloader:
          #  efi: { secureBoot: False}
        resources:
          requests:
            memory: 16Gi
            cpu: 4
      networks:
      - name: default
        pod: {}
      - multus:
          networkName: ovs-blue
        name: ovs-blue-net
      terminationGracePeriodSeconds: 0
      volumes:
      - name: datavolumedisk1
        persistentVolumeClaim:
          claimName: ubuntu-home
      - cloudInitNoCloud:
          userData: |-
            #cloud-config
            hostname: ubuntu-home
            fqdn: ubuntu-home.lan
            manage_etc_hosts: true
            users:
            - name: thiller
              sudo: ALL=(ALL) NOPASSWD:ALL
              groups: users, admin
              home: /home/thiller
              shell: /bin/bash
              lock_passwd: false
              ssh-authorized-keys:
              - ssh-rsa XXXXXXXXX
            # only cert auth via ssh (console access can still login)
            ssh_pwauth: false
            disable_root: false
            chpasswd:
              list: |
                thiller:xxxxx
              expire: False
          networkData: |
            version: 2
            ethernets:
              enp1s0:
                dhcp4: true
                dhcp6: true
              enp2s0:
                dhcp4: true
                dhcp6: true

        name: cloudinitdisk

th-2021 commented 3 years ago

The POD IP is assigned, but reverse address resolution as well as name resolution doesn't work.

thiller@tom-ml350:~/Projects/microk8s/kubevirt/ubuntu$ kubectl -n vm-images get vmi NAME AGE PHASE IP NODENAME READY ubuntu-home 2d6h Running 10.1.154.117 tom-ml350 True

$ kubectl -n vm-images exec -ti dnsutils -- nslookup ubuntu-home.vm Server: 10.152.183.10 Address: 10.152.183.10#53

** server can't find ubuntu-home.vm: NXDOMAIN

command terminated with exit code 1

kubectl -n vm-images exec -ti dnsutils -- nslookup 10.1.154.117 ** server can't find 117.154.1.10.in-addr.arpa: NXDOMAIN

command terminated with exit code 1

rhrazdil commented 3 years ago

Hello, As I understand it, setting the hostname via cloud-init doesn't propagate outside of the VM. Please, see this documentation section about DNS: https://kubevirt.io/user-guide/virtual_machines/dns/#dns-records

In a nutshell, you can set hostname and subdomain to the VM spec. I've tried the following scenario based on the documentation and your VM setup with bridge binding on pod network:

apiVersion: kubevirt.io/v1
kind: VirtualMachineInstance
metadata:
  labels:
    expose: me
  name: vmi-masquerade1
spec:
  hostname: "myvmi"
  subdomain: "mysubdomain"
  domain:
    devices:
      disks:
      - disk:
          bus: virtio
        name: containerdisk
      - disk:
          bus: virtio
        name: cloudinitdisk
      interfaces:
      - bridge: {}
        name: default
      rng: {}
    resources:
      requests:
        memory: 1024M
  networks:
  - name: default
    pod: {}
  terminationGracePeriodSeconds: 0
  volumes:
  - containerDisk:
      image: registry:5000/kubevirt/fedora-with-test-tooling-container-disk:devel
    name: containerdisk
  - cloudInitNoCloud:
      userData: |-
        #!/bin/bash
        echo "fedora" |passwd fedora --stdin
    name: cloudinitdisk
---
apiVersion: v1
kind: Service
metadata:
  name: mysubdomain
spec:
  selector:
    expose: me
  clusterIP: None
  ports:
  - name: foo # Actually, no port is needed.
    port: 1234
    targetPort: 1234

$ kubectl get pods -o wide
virt-launcher-vmi-masquerade1-vz8tc   2/2     Running   0          31m   10.244.196.146   node01   <none>           <none>

$ virtctl console vmi-masquerade1
ping myvmi.mysubdomain
PING myvmi.mysubdomain.default.svc.cluster.local (10.244.196.146) 56(84) bytes of data.
64 bytes from myvmi.mysubdomain.default.svc.cluster.local (10.244.196.146): icmp_seq=1 ttl=64 time=0.019 ms
64 bytes from myvmi.mysubdomain.default.svc.cluster.local (10.244.196.146): icmp_seq=2 ttl=64 time=0.028 ms
...

From another pod 
$ kubectl exec -it dnsutils -- /bin/sh
/ # nslookup myvmi.mysubdomain
Server:     10.96.0.10
Address:    10.96.0.10#53

Name:   myvmi.mysubdomain.default.svc.cluster.local
Address: 10.244.196.146

Hope this helps

th-2021 commented 3 years ago

I have set hostname and subdomain and it's not working. (Only with Debian and Ubuntu)

Am 26.10.2021 um 11:12 schrieb Radim Hrazdil:

Hello, As I understand it, setting the hostname via cloud-init doesn't propagate outside of the VM. Please, see this documentation section about DNS: https://kubevirt.io/user-guide/virtual_machines/dns/#dns-records

In a nutshell, you can set hostname and subdomain to the VM spec. I've tried the following scenario based on the documentation and your VM setup with bridge binding on pod network:

|apiVersion: kubevirt.io/v1 kind: VirtualMachineInstance metadata: labels: expose: me name: vmi-masquerade1 spec: hostname: "myvmi" subdomain: "mysubdomain" domain: devices: disks: - disk: bus: virtio name: containerdisk - disk: bus: virtio name: cloudinitdisk interfaces: - bridge: {} name: default rng: {} resources: requests: memory: 1024M networks: - name: default pod: {} terminationGracePeriodSeconds: 0 volumes: - containerDisk: image: registry:5000/kubevirt/fedora-with-test-tooling-container-disk:devel name: containerdisk - cloudInitNoCloud: userData: |- #!/bin/bash echo "fedora" |passwd fedora --stdin name: cloudinitdisk --- apiVersion: v1 kind: Service metadata: name: mysubdomain spec: selector: expose: me clusterIP: None ports: - name: foo # Actually, no port is needed. port: 1234 targetPort: 1234 |

$ kubectl get pods -o wide virt-launcher-vmi-masquerade1-vz8tc 2/2 Running 0 31m 10.244.196.146 node01

$ virtctl console vmi-masquerade1 ping myvmi.mysubdomain PING myvmi.mysubdomain.default.svc.cluster.local (10.244.196.146) 56(84) bytes of data. 64 bytes from myvmi.mysubdomain.default.svc.cluster.local (10.244.196.146): icmp_seq=1 ttl=64 time=0.019 ms 64 bytes from myvmi.mysubdomain.default.svc.cluster.local (10.244.196.146): icmp_seq=2 ttl=64 time=0.028 ms ...

From another pod $ kubectl exec -it dnsutils -- /bin/sh / # nslookup myvmi.mysubdomain Server: 10.96.0.10 Address: 10.96.0.10#53

Name: myvmi.mysubdomain.default.svc.cluster.local Address: 10.244.196.146

|Hope this helps |

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/kubevirt/kubevirt/issues/6675#issuecomment-951740553, or unsubscribe https://github.com/notifications/unsubscribe-auth/AVVFCF3P5AFG35WSUF2PJLDUIZ5IJANCNFSM5GVGUVDQ. Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

rhrazdil commented 3 years ago

Did you also create the service, that is called the same as the subdomain and is in the same namespace as the VM? The service also needs to have a selector matching the VMIs in the subdomain

th-2021 commented 3 years ago

Yes, I have the service, but forgot the label in the manifest. It's working now.

rhrazdil commented 3 years ago

/assign @rhrazdil

kubevirt / kubevirt

no dns entry for VM #6675