minikube start (kvm) "Error starting host: Error creating new host: dial tcp: missing address."

[broady]

Is this a BUG REPORT or FEATURE REQUEST? (choose one):
bug report

Minikube version (use minikube version):

minikube version: v0.14.0

Environment:

• OS (e.g. from /etc/os-release): ubuntu trusty
• VM Driver (e.g. cat ~/.minikube/machines/minikube/config.json | grep DriverName): kvm, but machines directory is empty
• Docker version (e.g. docker -v): 1.12.1
• Install tools:
• Others:

What happened:

$ minikube config set vm-driver kvm
$ minikube start                                                                                                                                                                                
Starting local Kubernetes cluster...
E1215 20:30:51.804680    3691 start.go:94] Error starting host: Error creating new host: dial tcp: missing address.

 Retrying.
E1215 20:30:51.805291    3691 start.go:100] Error starting host:  Error creating new host: dial tcp: missing address
Exit code 1

What you expected to happen:

How to reproduce it (as minimally and precisely as possible):

Anything else do we need to know:

[r2d4]

Hey we've actually seen this problem before and sent a PR upstream to the KVM driver that was merged, however I don't think theres been a recent release of it. (dhiltgen/docker-machine-kvm#18)

The fix is that you need to also install the libvirt-bin package on ubuntu

Ref #278

[broady]

$ sudo apt-get install libvirt-bin
Reading package lists... Done
Building dependency tree       
Reading state information... Done
libvirt-bin is already the newest version (1.2.2-0ubuntu13.1.17).
0 upgraded, 0 newly installed, 0 to remove and 1 not upgraded.

Did you mean I also need a new version of the kvm driver?

[r2d4]

No, that PR only made the error message clearer.

Have you added yourself to the libvirtd group?
You can verify that everything is running properly if you can run $ virsh -c qemu:///system list

[broady]

I did run sudo usermod -a -G libvirtd $(whoami)

$ cat /etc/group | grep libvir
libvirtd:x:134:cbro
$  virsh -c qemu:///system list
error: failed to connect to the hypervisor
error: Failed to connect socket to '/var/run/libvirt/libvirt-sock': Permission denied
Exit code 1

Oh, but looks like it works in a brand new tty (well, at least it looks like it's now doing something). D'oh. (but sweet!)

$ kubectl get nodes
NAME       STATUS    AGE
minikube   Ready     14s

[r2d4]

Closing this, looks like you got it working. Feel free to reopen if you're still having problems

[westlakem]

Any idea how we can fix this moving forward? i'm running on docker and I'm getting the same error "missing address"

[dlorenc]

Can you explain a little more what you're trying to do? This bug is about a permission issue when using the KVM driver.

[Anticom]

I'm on Ubuntu 16.04 and installed kubectl and minikube via Ubuntu's snap and I get the same errors:

$ minikube start --vm-driver="kvm"
There is a newer version of minikube available (v0.23.0).  Download it here:
https://github.com/kubernetes/minikube/releases/tag/v0.23.0
To disable this notification, add WantUpdateNotification: False to the json config file at /home/anticom/snap/minikube/4/.minikube/config
(you may have to create the file config.json in this folder if you have no previous configuration)
Starting local Kubernetes cluster...
E1115 02:25:58.211464    9757 start.go:83] Error starting host: Error creating new host: dial tcp: missing address. Retrying.
E1115 02:25:58.222801    9757 start.go:83] Error starting host: Error creating new host: dial tcp: missing address. Retrying.
E1115 02:25:58.232688    9757 start.go:83] Error starting host: Error creating new host: dial tcp: missing address. Retrying.
E1115 02:25:58.232710    9757 start.go:89] Error starting host:  Error creating new host: dial tcp: missing address
Error creating new host: dial tcp: missing address
Error creating new host: dial tcp: missing address

[gabrielmancini]

+1

[hudsantos]

+1

[vrajanap]

+1

[wskinner]

I also have this problem on Ubuntu 14.04.

[wskinner]

@r2d4 can you reopen this? Seems like a lot of people are having this issue now.

[zomglings]

+1

• minikube version: minikube version: v0.26.0
• OS: Ubuntu 16.04
• VM Driver: "DriverName": "kvm2",
• Docker version: Docker version 18.03.0-ce, build 0520e24

[oesgul]

+1

[razeone]

+1

• minikube version: v0.26.1
• OS: Fedora 27
• VM Driver: kvm/kvm2
• Docker version 18.04.0-ce, build 3d479c0

[jorgerod]

+1

minikube version: minikube version: v0.26.0
OS: Ubuntu 16.04
VM Driver: "DriverName": "kvm2",
Docker version: Docker version 18.03.0-ce, build 0520e24

[tjma2001]

I am having this issue with snap's version of minikube as well. I uninstalled the snap version and performed the installation with

curl -Lo minikube https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64 && chmod +x minikube && sudo mv minikube /usr/local/bin/

And the install seems to be proceeding without a hitch

[ostretsov]

Had the same issue. The problem was with my group. Make sure you are in libvirt (libvirtd group in 16.04) group:

$ id
uid=1000(artem) gid=133(libvirtd) groups=133(libvirtd),4(adm),24(cdrom),27(sudo),30(dip),46(plugdev),113(lpadmin),128(sambashare),999(docker),1000(artem)

[skis]

I have the same issue.
Virtual Machine - Ubuntu 18.10.
Tried KVM2 driver and yet. Anyone else experiencing this issue?

[vds]

+1

[Ddper]

+1

[Grigore147]

Same issue:
Ubuntu 16.04 with KVM2.
minikube version: v0.32.0

[alanwds]

Same here:

Ubuntu 14.04 with KVM2
minikube version: v0.32.0

[ugurcan377]

I had the same problem and the solution in this issue worked.
#3437

[TimZaman]

Still a problem! Reopen? Ubuntu 16.04 with KVM2.

Solution is to downgrade minikube to v0.30.0 if you are on 16.04

[alexellis]

I'm still getting this issue with Ubuntu 16.04 KVM2 and v0.33.1

[krilor]

I've just tried on Linux Mint 18 (based on Ubuntu 16.04) + KVM2 and got a similar error.

Downgraded to v0.30.0 helped.

[jvleminc]

Also happening on Ubuntu 18.04 after a fresh install of minikube.

The above workaround worked fine:

Uninstall snap version and install "real version" with

curl -Lo minikube https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64 && chmod +x minikube && sudo mv minikube /usr/local/bin/

[brennen]

Also happening on Ubuntu 18.04 after a fresh install of minikube.

Same; oddly I installed Minikube once on an existing Ubuntu system, then had to nuke and pave the machine for unrelated reasons and now find that I'm unable to get Minikube working again from the same instructions.

Edit: On double-checking myself, I find that the error message I'm seeing is instead:

Error starting host:  Error creating new host: Error attempting to get plugin server address for RPC: Failed to dial the plugin server in 10s

Apologies for noise in the likely event this is unrelated.

[kasvith]

Same issue, but using a new terminal or tty fixed issue for me

[chaitanyanettem]

For now on ubuntu 16.04, you can also get around this by installing virtualbox (sudo apt-get install virtualbox).

You can then start minikube with minikube start and it just works.

[sblcook]

This is still happening to me on minikube v1.1.0. Opening another terminal doesn't seem to help, but I am included in the libvirtd group.

OS: Ubuntu 16.04
VM Driver: kvm
Docker version: 18.06.0-ce,

[ngmariusz]

same here ( #4416 )

[kpooja20]

same problem after installing libvirt.

[afbjorklund]

For anyone else having the same issue, please check that you are in the group and that libraries are ok.

id should include "libvirtd", and for instance ldd should not show "version `LIBVIRT_3.0.0' not found"

Issue itself will soon be fixed in minikube 1.3.0 (#4555)

It was supposed to be fixed in minikube 1.0.0 (#3689)

[bezna-hack]

os: linux mint 19
minikube v1.6.2
vm driver: kvm
docker v19.03

I experienced the same issue after installing minikube with snap. For me the solution was to remove the snap install and delete the /snap/minikube folder and then install minikube wirh:

curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64
&& sudo install minikube-linux-amd64 /usr/local/bin/minikube

This is pretty odd since snap is suggested in the official documentation.
I also had to install docker-machine.

[MykytaLiubshin]

I am having this issue with snap's version of minikube as well. I uninstalled the snap version and performed the installation with

curl -Lo minikube https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64 && chmod +x minikube && sudo mv minikube /usr/local/bin/

And the install seems to be proceeding without a hitch

THANK YOU
FIXED IT FOR ME

[urbandroid]

uninstalling snap and installing "real" version is not a "real" fix. That makes you run a deceiving app unconfined, deceiving because even though they know their snap is broken they are acting like it is not and using as a honey pot.

[tjma2001]

I linked the official minikube binary. If that is a honey pot that I give up. It is running unconfined but I am not sure the snap version of minikube can run unconfined either.