-
Notifications
You must be signed in to change notification settings - Fork 657
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Multipass and Vagrant with VirtualBox clash on macOS #3674
Comments
Hi @pkarjala, Can you verify that your firewall is not enabled and that you are not running a VPN? |
I can confirm that I am not on a VPN, and that the macOS Firewall is disabled. Thanks! |
One of my colleagues wrote an excellent troubleshooting guide for ssh connection issues here From your posted logs I can see that the instance is booting up successfully, but for some reason Multipass is not able to connect to the instance. As explained in the above discussion you could try using tcpdump to check for dhcp requests and replies or check the contents of dhcp_leases. |
Thank you, I will review and follow up! |
Same issues. I have updated to macOS Sequoia today and multipass instance are not starting and are in an unknown state. |
@Gab2thebo Yes, there is an issue with new Sequoia updating causing issues with Multipass not being able to acquire IP addresses of instances. It is being tracked with #3661 |
Thank you again, I was able to get back to some troubleshooting on this issue following the steps listed in #3660 It looks like I am in situation 3, as each time an instance starts I can see the DHCP lease being granted, and it also shows up in
Of note, both of these happened after rebooting my system, and I was able to start and stop Multipass instances without issue. Then I started up one of my other VM services, and this is when it became an issue. I also run Vagrant machines on my system using VirtualBox as the VM host. This has not been an issue with conflicts with Multipass in the past. Running a Vagrant instance and then attempting to run a Multipass instance results in the Multipass instance being inaccessible. Doing the opposite results in both being able to run. But as long as any Vagrant instance is running, Multipass is unable to pull a dhcp lease from the host OS. The following is monitoring while Vagrant is running and attempting to boot a new Multipass instance, and then stopping Vagrant and running a Multipass instance:
So this appears to be tied to Vagrant (using VirtualBox as its VM host) conflicting with Multipass's ability to pull a DHCP lease for the booting instance. For now we are simply not running Vagrant while using Multipass, but will have to dig into why this is conflicting. |
Very interesting @pkarjala. So in the output of the last |
Also @pkarjala, what versions of Vagrant and VirtualBox are you using? |
Correct; this is after stopping all running Vagrant instances and then launching a Multipass instance. Additionally, it appears that when Multipass is running 1 or more instances and has bound to This example shows
Also, if it is relevant, this is running on an Intel Mac, not an Apple Silicon Mac. |
Thank you @pkarjala, I suspect this is down to how QEMU uses vmnet. But we can at least update the new troubleshooting guide we're working on. I understand the new network stack in VirtualBox 7.1.0 does not work for everyone, but if you ever upgrade to 7.1.X and have new findings, please share! |
I have no qualms with attempting to upgrade to VirtualBox 7.1.x (I honestly wasn't aware it was out until working on researching this issue); will likely try that early next week and report back. |
Had a few minutes and tested this morning, unfortunately it turns out the current version of Vagrant is not compatible (without some tweaking) with VirtualBox 7.1.x as per hashicorp/vagrant#13501 So will have to wait for a bit for proper testing. |
Describe the bug
When starting a new multipass instance, it will hang and the status becomes "Unknown".
To Reproduce
From either the GUI or the CLI, launch a new Ubuntu 24.04 LTS image with default settings.
multipass launch 24.04
Expected behavior
The instance should launch and be available for use.
Logs
When launching from GUI:
When launching from CLI:
Logs:
Additional info
multipass version
:multipass info
:multipass get local.driver
:qemu
Additional context
I have rebooted the host computer, as well as uninstalled and reinstalled multipass entirely. It is installed via the
.pkg
method (notbrew
).Initially an instance booted without issue. After that they all time out while attempting to boot. Stopping and restarting the initial working instance also resulted in a timeout and being unable to use it.
The text was updated successfully, but these errors were encountered: