As a starting point to help those with use of rvm and documentation for
development and contribution, ensure that some basic config to select a
ruby version and gemset to use is in placed.
Ensure that when no box is defined for a pxe install, that the domain
name is only populated for subsequent boots and if it is already
provided by a create action, skip attempting to retrieve from a
non-existing domain XML.
Fixes: #1508
Vagrant newer than 2.2.11 reworked how the after hook definition
functions requiring it to be called in a different way to ensure it is
possible to hook the box remove action to perform the expected local
action.
Add a compatibility function to ensure that the plugin works with both
mechanisms and some simple tests to ensure that with the unit tests
validated against the older vagrant versions that it confirms the action
will be called as expected.
As part of fixing the call of the remove_libvirt_image action, ensure it
only executes when the box removed is a libvirt one and ignores all
others.
Fixes: #1196
Basic tests for synced folder capabilities for 9p and virtiofs should
ensure consistent detection of files for coverage across multiple
versions of ruby as some versions pick up these files and others do not.
Detect when the device provided for connecting the VM to a public
network via the host does not exist and attempt to provide a useful
error message to give end users a hint of how to correct.
Fixes: #1477
Calling undefine on a domain and recreating it can result in some edge
case errors where if the current capabilities of libvirt have been
reduced, it may not be possible to restore the old definition.
Instead switch to calling `domain_define` with the new definition and
check that the resulting libvirt domain definition has been updated in
the expected manner, otherwise report an error to the user.
Fixes: #949
Relates-to: #1329
Relates-to: #1027
Relates-to: #1371
Builds on the work started by @randomvariable in #840 addressing the comments
left there.
Fixes: #529
Signed-off-by: Rodolfo Olivieri rodolfo.olivieri3@gmail.com
With multi volume boxes, need to ensure that disk settings such as the
device assigned are resolved dynamically once it has been established
which devices have already been assigned to the box volumes on either
initial creation or subsequent boots.
Otherwise users are forced to always explicitly define the device for
additional storage instead of having it be automatically assigned the
next available device.
Consequently previous changes have broken the ability for machines
with additional storage to be halted and restarted correctly.
Include an integration test that for additional storage checks that the
machine can be stopped and started again.
Fixes: #1490
When removing networks, skip calling destroy to deactivate the network
if it is not currently active. This avoids errors when the host has been
rebooted and vagrant machines are removed as the networks are typically
not automatically started.
The loader tag is required when nvram is being enabled. Ensure it is
marked as a requirement of validation and also support it's
configuration during the start domain action.
Switch to instance doubles for driver where possible and reduce usage of
allow_any_instance_of.
Additionally replace some incorrect allows with expects to validate
calls are not made.
Handle enabling and disabling of NVRAM during start domain.
This patch contains latent support for upstream PR fog-libvirt#102 to support destroy
with NVRAM enabled once that is merged.
Fixes#1027.
When assigning a pci passthrough device, the domain, bus, slot and
function parameters pertain to the host device address. Add guest_
flavours of these to allow setting the address of the device in the
guest as well.
Fixes: #1475
While trying to use the steps provided in the `Additional Notes for Fedora and
Similar Linux Distributions` I noticed that some of the versions referenced
there were not up-to-date, as well as a missing script call for the krb5.
Signed-off-by: Rodolfo Olivieri <rolivier@redhat.com>
Using `autoload` fairly consistently results in 'constant not defined'
errors for each of HandleBox, SyncedFolders, WaitForCommunicator
when using ruby 3.1. Presumably there's some sort of race condition.
Explicitly using `require` for these three makes the problem go away.
It might be a bit ugly to have those `require` lines right at the
point where those symbols are used, vs. the neat `autoload` section
at the top of the file, but I'm not sure how better to structure this.
Suggestions most welcome :-)
Fixes: #1445
Signed-off-by: Tim Serong <tserong@suse.com>
Provisioning is executed on the out-call action, so it should be added
the call start before the configuration of the network interfaces is
scheduled via a guest capability as it is also executed out the out call
to allow the first part to be performed before the VM is booted and the
latter to be done after it is running.
Fixes: #1479
Avoid crash caused by power management suspension where fog fails to translate domain state pmsuspended.
Changed the driver.state() function so that it does not crash on undefined domain state. Just return nil :unknown instead.
Added unittests and made existing more general.
Closes#1472
Switch to the command as an array format when passing the port
forwarding ssh command to spawn to switch to execution of the ssh
process directly instead of being executed by the default shell.
Fixes: #1467
Bundler for older ruby versions uses anonymous git to clone ruby
dependencies directly from repos. Ensure these all use https by
configuring git to replace the URL automatically.
When running the docker container, only output what is returned by
vagrant to stdout, and redirect everything else to stderr to facilitate
anything parsing stdout.
Fixes: #1463
Facilitate disabling serial devices by allowing for the configuration to
explicitly set serials to an empty array and ignoring adding the default
serial if this is the case.
Autoloading interaction with threads can cause issues. Follow the best
practice of performing the autoloading early in the hope that this
avoids the issue with a symbol not being fully defined before
another thread attempts to use the symbol in the same file.
Fixes: #1445
Trying to create a routed network with just ipv6 fails with:
Error occurred while creating new network: Call to virNetworkDefineXML failed: XML error: Missing required address attribute in network 'test'.
Since an ipv4 address should be optional for this scenario, make it so.
This makes the following configuration work:
config.vm.network :private_network,
:libvirt__network_name => "test",
:libvirt__forward_mode => "route",
:libvirt__dhcp_enabled => false,
:libvirt__guest_ipv6 => "yes",
:libvirt__ipv6_address => "fc00::",
:libvirt__ipv6_prefix => "64"
When the :ip is ipv6, set the 'static6' type, so that we select the
correct _static6.erb template to configure the interface in the guest.
With the following configuration, ipv6 works out of the box on CentOS 8:
(1..2).each do |i|
config.vm.define "node#{i}" do |node|
node.vm.hostname = "node#{i}"
node.vm.network :private_network,
:ip => "fc00::#{i+9}",
:netmask => "64",
:libvirt__dhcp_enabled => false,
:libvirt__forward_mode => "veryisolated",
:libvirt__network_name => "test"
end
end
Fixes: #914
Libvirt did already change the current 9216kb
to 16384kb when starting a domain so in practice this is a no-op.
The libvirt default was changed in 2014 in 81ba2298b2
Starting from version 0.7.0 (https://newreleases.io/project/github/vagrant-libvirt/vagrant-libvirt/release/0.7.0), always_destroy is set to true by default for any network.
However, in certain cases, it would be useful to keep the management network, especially when it has been already configured (externally) with Static IP leases.
This option allows to change the always_destroy behaviour for the management network.
Reduce the patching needed should a distro wish to switch the default
from using the system connection by default to using a session
connection by default.
Should now only require patching the default value and a single test
checking the defaults.
When cpu mode was set to host-passthrough the template would skip
setting all other settings for the cpu, while it appears from the
documentation that it supports use of feature elements, and testing
confirms that it also supports the topology element.
https://libvirt.org/formatdomain.html#cpu-model-and-topologyFixes: #975
Adjust handling to use a method reference instead of attempting to
update the env value as that only appears to be available to actions
further down the chain and any changes are not visible during recovery
when exercised for real.
By passing a function reference can limit the actions that can be taken
to only allow a single call that will set a flag internal to the clean
up action.
Fixes: #1328
Switch to dedicated actions to trigger destroy/halt via the recovery
mechanism if the VM bring up has failed to complete the bring up of the
machine.
Fixes: #1328
Switch to using explicit references to objects to be partially mocked
and remove the need to resolve the string constants as this will catch
more instances of calls to invalid or missing methods.
Rework how the vm is added to the machine for one of the tests as it is
not a method and instead is provided via internal state being exposed
with a helper.
When the tests are executed and rsync is not installed on the machine executing
the tests, then vagrant-libvirt will automatically fallback to virtiofs or
9p. Both of these perform the following call to check the libvirt version:
libvirt_version = machine.provider.driver.connection.client.libversion
However, this mock was not setup and thus those tests would fail.
This fixes https://github.com/vagrant-libvirt/vagrant-libvirt/issues/1415
In PrepareNFSSettings.read_host_ip we use a UDPSocket to connect to the host in
question. However, this fails completely if the system executing the tests has
no network adapters enable, like builders in OBS or Koji. If we double the calls
to this, then we no longer face that issue.
This requires that the qemu_use_agent is enabled and at least one
network is defined in the Vagrantfile that is either private or public.
The restriction that qemu_use_agent need be enabled may be removed
subsequently for networks using a static IP address. That should be
considered a subsequent problem.
Fixes: #1204Fixes: #1124Fixes: #1057
