- Suspend/Resume: replace fixed sleep with polling loop (up to 20s)
- Restart: ensure VM is running (not paused) before attempting restart
- SDN: add propagation delays between subnet/vnet/zone removal
- OVA: split into OVF parsing test (always works) and import test
(tolerates synthetic VMDK failures gracefully)
- Dockerfile: add qemu-utils for proper VMDK generation
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Suspend/Resume: move to Linux VM lifecycle context (empty VMs may
not reliably transition to paused state)
- SDN subnet model: dhcp-range is an array, not string
- SDN subnet delete: use PVE-format ID (vnet-ip-prefix) not CIDR
- Container config: trim trailing newline from PVE description
- OVA: use qemu-img to create valid sparse VMDK (add qemu-utils to
test container Dockerfile)
- Enable 'images' content type on local storage for OVA import
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- SDN: zone ID cannot contain hyphens — rename 'pester-zone' to 'pesterz'
- SDN: subnet removal uses original CIDR, not transformed ID
- Containers: New-PveContainer now auto-assigns VM ID via /cluster/nextid
when -VmId is omitted (was missing, unlike New-PveVm)
- OVA: enable 'images' content type on local storage for import-from
- Suspend: add delay before status check, use Should -Not -Throw pattern
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
PVE upload API validates file extensions per content type. The
content=import type does not accept .img extension. The Ubuntu cloud
image is qcow2 format, so rename from .img to .qcow2 during download.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Send-PveFile replaces Send-PveIso with a -ContentType parameter
(iso, vztmpl, import). This is a breaking rename — the module has
not been released yet.
Test infrastructure refactored:
- prepare-test-vm.sh replaced with prepare-test-environment.sh
(only SSH ops: enable content types, upload cloud-init snippet,
curl download cloud image)
- VM provisioning moved into integration tests as "Linux VM —
Provisioning" context: Send-PveFile, New-PveVm, Import-PveVmDisk,
Set-PveVmConfig, Set-PveCloudInitConfig, Start-PveVm,
Test-PveVmGuestAgent — all tested as part of the test suite
- PVETEST_LINUX_VMID env var removed (VM created by tests)
- PVETEST_CLOUD_IMAGE_PATH + PVETEST_PASSWORD added
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Absolute filesystem paths are rejected by PVE API when using API token
auth (only root ticket auth allows arbitrary paths). Instead:
- Enable 'import' content type on local storage alongside snippets
- Download cloud image with content=import (not iso)
- Reference as local:import/filename in Import-PveVmDisk
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The cloud image is downloaded as content=iso, but import-from requires
content type 'images' or 'import'. Use the absolute node path
(/var/lib/vz/template/iso/) instead of the storage volume reference
(local:iso/) to work around this constraint.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Parses OVF metadata client-side from the OVA TAR archive to extract
VM name, CPU, memory, disks, and network adapters. Then uploads the
OVA with content=import, creates the VM, and imports each disk.
- OvfMetadata parser with multi-target TAR support (System.Formats.Tar
on net9.0, manual 512-byte header parsing on netstandard2.0/net48)
- VmService.UploadOva() for content=import uploads
- Import-PveOva cmdlet with -Name/-Memory/-Cores overrides and -Wait
- Pester unit tests
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Imports disk images (qcow2, raw, vmdk) and OVA archives into VMs
via the PVE API using POST config with import-from syntax.
- VmService.ImportDisk() method using POST /nodes/{node}/qemu/{vmid}/config
- Import-PveVmDisk cmdlet with -Source, -Disk, -TargetStorage, -Format, -Wait
- Pester unit tests
- prepare-test-vm.sh updated to use Import-PveVmDisk instead of SSH qm importdisk
- Set-PveNetwork: add required Type parameter (fixes PVE API 400 error)
- Integration test fix: pass -Type bridge to Set-PveNetwork
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
ConfirmImpact.High causes interactive confirmation prompts in
non-interactive CI, blocking before ProcessRecord runs. Add
-Confirm:$false to all Stop-PveVm, Reset-PveVm, Remove-*,
Restore-*, and New-PveTemplate calls in both unit and integration tests.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- New Move-PveContainer cmdlet with -Online and -Wait support
- Uses existing ContainerService.MigrateContainer
- Pester unit tests added to ContainerLifecycle.Tests.ps1
- README cmdlet reference updated
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Add bridge create/update/remove/apply test cases
- Add interface type filter test
- Tests exercise New-PveNetwork, Set-PveNetwork, Remove-PveNetwork,
and Invoke-PveNetworkApply against live PVE
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Container snapshots (4 cmdlets):
- Get-PveContainerSnapshot, New-PveContainerSnapshot
- Remove-PveContainerSnapshot, Restore-PveContainerSnapshot
- Service methods in ContainerService
- Pester unit tests
SDN subnet management (3 cmdlets):
- Get-PveSdnSubnet, New-PveSdnSubnet, Remove-PveSdnSubnet
- PveSdnSubnet model and NetworkService methods
- Pester unit tests
Also updates manifest (73 cmdlets), README cmdlet reference, and CHANGELOG.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
PVE 9 may not immediately reflect the template flag in the VM list
after New-PveTemplate returns. Add a 3-second delay before verifying.
PVE 8 passed without the delay.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
ConfirmImpact.High causes an interactive prompt in CI without
-Confirm:\$false, leading to a null reference error.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The cmdlet calls the /cloudinit/dump endpoint which returns YAML
content, not a task UPID. This gets passed to WaitForTask causing a
501 error. PVE regenerates cloud-init automatically on VM start, so
the explicit regenerate call is not needed for test VM setup.
Bug to fix: RegenerateCloudInitImage should PUT to /cloudinit instead
of GET from /cloudinit/dump.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The Debian .qcow2 extension was rejected by PVE's download-url API.
Ubuntu's noble-server-cloudimg-amd64.img has a .img extension that
PVE accepts, and cloud-init + qemu-guest-agent work the same way.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace raw qm/SSH commands with module cmdlets where possible:
- Invoke-PveStorageDownload for cloud image download
- New-PveVm for VM creation
- Set-PveCloudInitConfig for cloud-init user/password/IP
- Invoke-PveCloudInitRegenerate to apply config
- Start-PveVm to boot the VM
SSH/SCP retained only for operations without API support:
- Snippet upload (no snippet API)
- pvesm set for snippets content type
- qm importdisk (no importdisk cmdlet)
- qm set for --scsi0, --cicustom, --agent, --boot (not in Set-PveVmConfig)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Replace virt-customize with a cloud-init custom user-data snippet that
installs qemu-guest-agent on first boot. The snippet is SCP'd to the
nested PVE's /var/lib/vz/snippets/ and referenced via --cicustom.
This eliminates the need for libguestfs-tools entirely — no install
on nested PVE, no extra Docker image, no large dependencies.
Timeout increased to 300s to allow for cloud-init package installation.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- prepare-test-vm.sh: install libguestfs-tools if not present on the
fresh nested PVE before running virt-customize
- Mask PVE_PASSWORD in logs via ::add-mask:: to prevent leaking
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Move virt-customize to run on the nested PVE host (which already has
libguestfs) instead of inside the CI container. This removes
libguestfs-tools, linux-image-generic, and qemu-utils from the
Dockerfile, shrinking the image from ~1.2GB to ~770MB.
- prepare-test-vm.sh: downloads Debian cloud image on nested PVE,
runs virt-customize there to install qemu-guest-agent, then creates
VM with imported disk + cloud-init
- Remove build-test-image.sh (no longer needed)
- Remove /opt/pve-images volume mount
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
proxmox-auto-install-assistant requires xorriso for ISO manipulation.
Expand the verification step to check all tools the scripts depend on.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Infrastructure:
- build-test-image.sh: Downloads Alpine cloud image, uses virt-customize
to install qemu-guest-agent, caches at /opt/pve-images/
- prepare-test-vm.sh: SCPs image to nested PVE, creates VM via
qm importdisk with cloud-init and agent enabled, waits for agent
Workflow:
- Build Alpine image step (cached on runner, only built once)
- Deploy test VM step after provisioning, passes LINUX_VMID to tests
Tests:
- Guest Agent VM lifecycle: verify running, ACPI restart, ACPI stop
- Templates: convert Linux VM to template, clone from template
(New-PveTemplate, New-PveVmFromTemplate)
- Fix Get-PveTask: use .Upid property instead of whole PveTask object
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
PVE 8 (Bookworm) may not have /etc/apt/sources.list.d/debian.sources,
causing the grep fallback to use 'trixie' which is invalid for PVE 8.
Use /etc/os-release VERSION_CODENAME which works on both PVE 8 and 9.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- New-PveRole: Privileges is a comma-separated string, not an array
- Set-PvePermission: uses UgId+Type, not Users
- Get-PveVmConfig: no Name property — test config get/set together
- Restart-PveVm: replace with Reset-PveVm (ACPI timeout on no-OS VM)
- Remove resize disk test (Set-PveVmConfig lacks Settings param)
- Copy-PveVm: must provide NewVmId explicitly
- Get-PveTask: requires Upid — test via Start-PveVm task chain
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Workflow changes:
- Convert to matrix strategy testing PVE 8.4 and 9.1 sequentially
- Each version gets unique VMID (99908/99909) and VM name
- Pass PVETEST_PVE_VERSION env var for version-specific assertions
- Version-specific artifact names for test results
New integration tests:
- User CRUD: create, list, update (Set-PveUser), remove
- API Token CRUD: create, list, remove (on dedicated test user)
- Role CRUD: create with privileges, list, remove
- Permissions: list and set (Get/Set-PvePermission)
- VM config: get and set config, resize disk
- VM lifecycle: restart, explicit remove
- Snapshots: full lifecycle (create, list, restore, remove)
- Storage: list content (Get-PveStorageContent)
- Tasks: list recent tasks (Get-PveTask)
- Version assertion: verify PVE major version matches expected
Coverage improved from 15 to ~30 cmdlets tested against live API.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
PS 5.1 cannot parse the ?? operator (PS 7.0+), causing a discovery
error even though integration tests are excluded by tag. Replace with
if/else that works on all PS versions.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
PS 5.1 prompts interactively for mandatory parameters instead of
throwing, causing CI to hang indefinitely. Replace invocation-based
tests with attribute-based IsMandatory checks where duplicates exist,
or convert to attribute checks where they don't.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The net9.0 DLL references System.Runtime 9.0.0.0 which is unavailable
on GitHub runners where pwsh ships with .NET 8.x. Adding netstandard2.0
as a target produces a DLL compatible with any .NET version (PS 5.1 on
.NET Framework through PS 7.x on .NET 8/9).
- Add netstandard2.0 to TargetFrameworks in Core and module csproj
- Update #if NET48 conditionals to include NETSTANDARD2_0
- Switch CI workflows from net9.0 to netstandard2.0 for Pester tests
- Update _TestHelper.ps1 framework search order
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Previous approach of env: DOTNET_ROOT='' at the step level wasn't
sufficient — the .NET runtime may have already cached the original
DOTNET_ROOT before pwsh started. Now:
- Remove *.deps.json AND *.runtimeconfig.json from publish output
- Unset DOTNET_ROOT via $env:DOTNET_ROOT = $null inside the pwsh script
- _TestHelper.ps1 also unsets DOTNET_ROOT and cleans deps/runtimeconfig
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
All CI now runs on the self-hosted runner using Docker containers,
eliminating GitHub-hosted runner minute consumption:
- Unit tests: PS 7.5 container (mcr.microsoft.com/powershell:7.5-ubuntu-24.04)
with .NET SDK 9.0 installed at build time
- Build/xUnit: .NET SDK 9.0 container (mcr.microsoft.com/dotnet/sdk:9.0)
- Integration tests: unchanged (runs directly on self-hosted runner)
Drops net48/PS 5.1 CI matrix (Windows-only, can't run in Linux
containers). Local Docker testing available via tests/Dockerfile.test.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Loading via the module manifest (.psd1) ensures PowerShell sets up
proper assembly resolution context. Loading via raw DLL path bypasses
PS's module loading infrastructure and fails on CI runners where the
.NET SDK's assembly resolution conflicts with PS's bundled runtime.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The deps.json file causes .NET's AssemblyLoadContext to conflict with
PowerShell's own runtime assembly resolution, resulting in
FileNotFoundException for System.Runtime 9.0.0.0 on all PS 7.x
platforms. Binary PS modules should not include deps.json.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- CS8601: Add null-forgiving operator on guarded dictionary assignments
in cmdlets where IsNullOrEmpty check precedes the assignment
- CS8602: Add Assert.NotNull after JObject["data"] in xUnit model tests
(JToken indexer returns nullable on net48)
- CS8604: Add null-forgiving on guarded arguments in NodeService,
WaitPveTaskCmdlet, GetPveTemplateCmdlet, and auth test parameters
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Coverage for 16 cmdlets that had no unit tests:
- Nodes: Get-PveNode, Get-PveNodeStatus
- VM config: Get-PveVmConfig, Set-PveVmConfig, Resize-PveVmDisk
- VM ops: Copy-PveVm, Move-PveVm
- Container config: Copy-PveContainer, Get-PveContainerConfig,
Set-PveContainerConfig
- Storage: Invoke-PveStorageDownload
- Roles: New-PveRole, Remove-PveRole
- Tasks: Get-PveTask, Wait-PveTask
Each test covers: command existence, parameter metadata, pipeline
support, ShouldProcess/WhatIf, and no-session error handling.
Total Pester unit tests: 705 (was 431)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- CS8604: Use cmdlet's own Node property instead of nullable task.Node
in WaitForTask calls across all VM cmdlets
- CS8618: Make PveSession.ServerVersion nullable (PveVersion?) since it
is set post-construction by PveAuthenticator
- CS8603: Add null-forgiving operator where IsNullOrWhiteSpace guards
guarantee non-null in PveHttpClient
- CS1573: Add XML param documentation to all public service methods
- xUnit1012: Make test parameters nullable (string?) for null test cases
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Create shared _TestHelper.ps1 for reliable module loading in both
local dev and CI (fixes System.Runtime 9.0.0.0 FileNotFoundException
on PS 7.x by trying Import-Module by name first, then local paths)
- Use (Get-Module).ModuleBase for manifest path resolution
- Remove PSProxmoxVE.MockServer project and MockIntegration tests
- Remove MockIntegration from ExcludeTag filters
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Change test projects from net10.0 to net9.0 to match source projects
- Update build workflow from net8.0 to net9.0
- Upgrade checkout to v5, setup-dotnet to v5 across all workflows
- Remove outdated PS 7.2 test matrix entries
- Restructure integration tests: create test keeps VM alive so start/stop,
snapshot, and cloud-init tests can use it instead of skipping. AfterAll
handles cleanup.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Change FQDN from pve-test.local to pve.test.local so the PVE node
name is "pve" (matching what tests expect)
- Mask the dynamically created API token and VM IP with ::add-mask::
before they appear in workflow logs
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
/tmp on the self-hosted runner has sticky bit — files created by
previous runs (or different users) can't be overwritten. RUNNER_TEMP
is per-run and owned by the runner user.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The proxmox-auto-install-assistant defaults --tmp to the source ISO's
directory (/opt/pve-isos/), which the runner user can't write to.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- first-boot.sh: Remove .sources files (PVE 9 DEB822 format) not just
.list, and detect suite (trixie vs bookworm) from debian.sources
- create-api-token.sh: Accept full endpoint URL (fixes missing port),
use /access/domains for API check (no auth needed on PVE 9),
avoid pipefail on curl|python3 pipe
- preflight-cleanup.sh: New script that runs at start and end of every
provisioning run — cleans orphaned VMs, ISOs, and stale terraform state
- Workflow: Add pre-flight and final cleanup steps so failed runs can't
leave the runner/host in a broken state
All changes validated manually on the self-hosted runner.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
OVMF was rebooting back into the installer CD after PVE install completed.
Setting boot_order to scsi0 first means the first boot falls through to
CD (empty disk), and after install OVMF boots from the EFI entry on disk.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Use proxmox_virtual_environment_file resource to upload the auto-install
ISO via the API. terraform destroy now cleans up both the VM and ISO.
- Remove manual upload-to-pve.sh and cleanup-pve-storage.sh scripts.
- Commit .terraform.lock.hcl for reproducible CI builds.
- Replace iso_file_id variable with iso_local_path and iso_storage.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Clint Branham