Implement LHv2 volume expansion

[albinsun]

Which issue(s) this PR fixes:

• Issue [e2e] [FEATURE] Support Longhorn V2 volume expansion #2178
• Issue [TEST][Infra] Fix common teardown in robot test #2672

What this PR does / why we need it:

Feature

1. Implement LHv2 volume expansion

Enhancement

1. Use NVMe disk instead of just large disks
2. Introduce run strategy coupling on blockdevice and setting components
3. Add new component storageclass
4. Add Common Suite/Test Teardown
   • Cleanup resources only when all tests Passed -> Keep env. for investigation (Resources will be cleanup in following run if tests are Pass)
   • Log Variables when tests Failed -> For Debugging
5. Adjust arguments of VM is created

Fix

1. Fix common_keywords.cleanup_volumes in Cleanup test resources

Special notes for your reviewer:

Additional documentation or context

Verification

• log.html
  
• report.html
  

Why Use NVMe disk instead of just large disks, NVMe vs. SAS

• NVMe

  2026-05-29.12.23.25.mov

• SAS

  2026-05-29.09.38.17.mov

[Copilot]

Pull request overview

Adds end-to-end Robot Framework coverage for Longhorn v2 volume expansion (issue #2178), introduces a reusable storageclass component (CRD/REST layered like the other components), and reworks NVMe disk selection plus shared suite/test teardowns. As part of the change, the positional argument order of VM is created / vm.create(...) was swapped (image_id moved before cpu/memory) and the existing test_vm.robot caller was updated accordingly. The setting and blockdevice components are refactored from an env-var-driven strategy selector to a CRD-first / REST-fallback pattern via NotImplementedError.

Changes:

• New test_longhorn_v2.robot flow covering provisioning, SC creation, VM-with-LHv2-volume, volume expansion (PVC-direct and VM-annotation-based), with NVMe disk picker and tag-based disk selector.
• New libs/storageclass/ component + vm.update_disk_size / volume keyword resource for expansion, plus Longhorn-node disk tag helpers in host.
• API-shape change: vm.create(...) signature reordered to (vm_name, image_id, cpu, memory, **kwargs); volume get_status renames state → phase and adds conditions.

Reviewed changes

Copilot reviewed 37 out of 37 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
harvester_robot_tests/tests/regression/test_longhorn_v2.robot	New LHv2 provisioning/expansion test flow and suite setup/teardown.
harvester_robot_tests/tests/regression/test_vm.robot	Updates caller to new VM is created positional arg order.
harvester_robot_tests/libs/vm/{base,vm,rest,crd}.py	Reorders create(...) args; adds update_disk_size + extra-disk support in CRD VM spec.
harvester_robot_tests/libs/volume/crd.py	Returns phase/conditions from get_status (renames state).
harvester_robot_tests/libs/storageclass/{init,base,crd,rest,storageclass}.py	New StorageClass component with CRD-first/REST-fallback pattern.
harvester_robot_tests/libs/setting/{base,crd,rest,setting}.py	Refactors to CRD-first/REST-fallback via NotImplementedError.
harvester_robot_tests/libs/blockdevice/{base,crd,rest,blockdevice}.py	Same CRD-first/REST-fallback refactor.
harvester_robot_tests/libs/host/{base,crd,host,rest}.py	Adds add/remove_lh_node_disk_tag and switches to CRD-first/REST-fallback base.
harvester_robot_tests/libs/keywords/{vm,volume,storage,host,common}_keywords.py	Exposes new component operations and get_timestamp.
harvester_robot_tests/libs/utility/utility.py	Adds get_timestamp helper.
harvester_robot_tests/keywords/{volume,virtualmachine,storage,image,host,common,variables}.resource	New volume keywords, refactored storage/SC keywords, NVMe disk picker, Common Suite/Test Teardown, env-overridable WAIT_TIMEOUT.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[asc5543]

lgtm

[khushboo-rancher]

Thanks @albinsun for the PR, looks good. Few comments.

[albinsun]

Changes

• Move storage class keywords to a independent storageclass.resource
• Single layer Suite Setup

Test Report

[khushboo-rancher]

Looks good, few minor comments

[khushboo-rancher]

If the LHv2 already enabled, how is it handled? Will it fail at 'setting.LHv2 Data Engine Is Disabled'?
Should we handle this while enabling of v2 data engine, if already enabled -> pass otherwise enable?

[albinsun]

Addressed in commit 60cfed6

[khushboo-rancher]

IMO, let's keep the 'storage' tag. storage is the group where we can keep all the storage related test cases like v2, lvm etc

[albinsun]

Addressed in commit 60cfed6

[khushboo-rancher]

This, Create LHv2 Storage Class & Create VM With LHv2 Volume should be P0. Other tests like Expand LHv2 Volume Size should be P1.

We can follow below guideline:
P0: The core "happy path" or main workflow of the feature.
Impact: If this fails, the feature is completely unusable, and there is no workaround.

P1: Secondary paths, edge cases, or essential supporting functionality.
Impact: The core feature works, but key high-value additions or sub-features are broken.

P2: "Nice-to-have" features, or rarely used workflows.
Impact: Low user impact; the core value of the feature remains fully intact.

[albinsun]

Addressed in commit 60cfed6

[albinsun]

Changes

1. Skip enabling Lhv2 data-engine if it's already enabled
2. Keep storage tag
3. Adjust test case priority

Report