### Install envtest Binaries

Source: https://github.com/cozystack/blockstor/blob/main/tests/integration/README.md

Installs the necessary `envtest` binaries for running integration tests. This is a one-time setup step.

```bash
go install sigs.k8s.io/controller-runtime/tools/setup-envtest@latest
export KUBEBUILDER_ASSETS="$(setup-envtest use --print path 1.34.x)"
```

--------------------------------

### Setup and Cluster Shape Example

Source: https://github.com/cozystack/blockstor/blob/main/tests/e2e/cli-matrix/README.md

This snippet illustrates the initial steps within a test cell: sourcing lib.sh, setting up the linstor CLI, and defining a cluster shape using kubectl or linstor commands.

```shell
#!/bin/bash

# Source the common library
. lib.sh

# Setup linstor CLI and build LCTL array
linstor_cli_setup

# Lay down a cluster shape (e.g., 2 nodes, 2 replicas)
# Example: kubectl apply -f cluster-shape.yaml
# Or chaining linstor commands:
# linstor rd c --name node-1 --role Controller
# linstor vd c --name vol-1 --size 10GiB --pool-name pool-1
# linstor r c --name vol-1 --node node-1 --pool-name pool-1 --role Controller
```

--------------------------------

### Kernel Probe Example

Source: https://github.com/cozystack/blockstor/blob/main/tests/e2e/cli-matrix/README.md

This example demonstrates how to check the kernel-side state of a DRBD resource on a specific node using drbdsetup.

```shell
# Check kernel-side DRBD status
on_node node-1 drbdsetup status my-resource
```

--------------------------------

### Install Linstor CLI

Source: https://github.com/cozystack/blockstor/blob/main/tests/integration/README.md

Installs the upstream `linstor` CLI and its Python dependencies, required for interacting with the storage system during integration tests. This is a one-time setup step.

```bash
apt-get install -y linstor-client python3-linstor   # Debian/Ubuntu
```

--------------------------------

### Resource Status Check Example

Source: https://github.com/cozystack/blockstor/blob/main/tests/e2e/cli-matrix/README.md

This example shows how to retrieve the observer-stamped status of a resource on a specific node using kubectl.

```shell
# Get resource status
kubectl get resource my-resource.node-1 -o json
```

--------------------------------

### Setup and Run Controller Integration Tests

Source: https://github.com/cozystack/blockstor/blob/main/tests/scenarios/README.md

Set up the environment for controller integration tests and then run them. This involves setting up envtest.

```bash
make setup-envtest && go test ./internal/controller/...
```

--------------------------------

### LINSTOR Resource Configuration Example

Source: https://github.com/cozystack/blockstor/blob/main/docs/design-res-rendering-linstor-parity.md

This is an example of a resource configuration file generated by LINSTOR. It includes settings for options, network, volumes, and connections.

```plaintext
resource "pvc-..."
{
    options {
        on-no-data-accessible suspend-io;
        on-no-quorum suspend-io;
        on-suspended-primary-outdated force-secondary;
        quorum majority;
    }

    net {
        cram-hmac-alg     sha1;
        shared-secret     "lFvd6mblDkeOmjc2YxEG";
        connect-int       15;
        ping-int          15;
        ping-timeout      20;
        protocol          C;
        rr-conflict       retry-connect;
        timeout           100;
        verify-alg        "crc32c";
    }

    on "node1" {
        volume 0 {
            disk      /dev/zvol/data/pvc-..._00000;
            disk { discard-zeroes-if-aligned yes; rs-discard-granularity 16384; }
            meta-disk internal;
            device    minor 1040;
        }
        node-id 0;
    }

    on "node0" {
        volume 0 {
            disk      none;                                      # diskless replica
            disk { discard-zeroes-if-aligned yes; rs-discard-granularity 16384; }
            meta-disk internal;
            device    minor 1040;
        }
        node-id 2;
    }

    on "node2" {
        volume 0 {
            disk      /dev/drbd/this/is/not/used;                # tiebreaker placeholder
            ...
        }
        node-id 1;
    }

    connection {
        net { connect-int 15; ping-int 15; ... }
        host "node1" address ipv4 10.4.100.12:7035;
        host "node0" address ipv4 10.4.100.11:7035;
    }
    connection {
        net { ... }
        host "node1" address ipv4 10.4.100.12:7035;
        host "node2" address ipv4 10.4.100.13:7035;
    }
}

```

--------------------------------

### ZFS Create Command for Thin Provisioning

Source: https://github.com/cozystack/blockstor/blob/main/docs/zfs-thick-validation-2026-05-14.md

Example of the ZFS `create` command used for thin provisioning, which includes the `-s` (sparse) flag.

```bash
zfs create -V N -s
```

--------------------------------

### ZFS Create Command for Thick Provisioning

Source: https://github.com/cozystack/blockstor/blob/main/docs/zfs-thick-validation-2026-05-14.md

Example of the ZFS `create` command used for thick provisioning, which does not use the `-s` (sparse) flag.

```bash
zfs create -V N
```

--------------------------------

### Example of Test-Only LUKS Layer CLI Cells

Source: https://github.com/cozystack/blockstor/blob/main/docs/cli-parity-audit-2026-05-19-refresh.md

This refers to command-line interface cells used specifically for testing the LUKS (Linux Unified Key Setup) layer. These are not intended for production use.

```bash
LUKS layer operator-CLI cells
```

--------------------------------

### Example CLI Command for Listing Resources

Source: https://github.com/cozystack/blockstor/blob/main/docs/cli-parity-audit-2026-05-19-refresh.md

This command lists resources. It was used in an audit to check the post-state of resources, but did not fully capture the behavior of resource creation with specific flags.

```bash
linstor r l
```

--------------------------------

### Example CLI Command for Creating a Resource

Source: https://github.com/cozystack/blockstor/blob/main/docs/cli-parity-audit-2026-05-19-refresh.md

This command creates a resource. It was identified that omitting the `--diskless` flag could lead to incorrect resource states due to skipped pool resolution.

```bash
linstor r c <node> <rd>
```

--------------------------------

### DRBD Net Block Configuration for Verify Algorithm

Source: https://github.com/cozystack/blockstor/blob/main/docs/design-res-rendering-linstor-parity.md

Example of how the verify-alg option is rendered within a DRBD net block for integrity checking.

```text
net {
    ...
    verify-alg "crc32c";
}
```

--------------------------------

### Example CLI Command for Volume Creation

Source: https://github.com/cozystack/blockstor/blob/main/docs/cli-parity-audit-2026-05-19-refresh.md

This command creates volumes for a resource. A bug occurred when late volume creation after the resource reached 'UpToDate' state resulted in new volumes being diskless.

```bash
vd c
```

--------------------------------

### Deploy and Run End-to-End Tests

Source: https://github.com/cozystack/blockstor/blob/main/tests/scenarios/README.md

Deploy the Blockstor stack and run end-to-end tests for a specific scenario. This involves multiple make targets for setup.

```bash
# E2E (full stand):
make up NAME=e2e7 && make piraeus NAME=e2e7 && make blockstor NAME=e2e7 && make pools NAME=e2e7
make e2e NAME=e2e7 SCENARIO=<scenario-script-name>
```

--------------------------------

### Example CLI Command for Diskless Resource Update

Source: https://github.com/cozystack/blockstor/blob/main/docs/cli-parity-audit-2026-05-19-refresh.md

This command updates a resource to be diskless. A bug caused it to return success even when the DRBD kernel did not detach the disk.

```bash
r td --diskless
```

--------------------------------

### Example CLI Command for Multi-Place Autoplace

Source: https://github.com/cozystack/blockstor/blob/main/docs/cli-parity-audit-2026-05-19-refresh.md

This command attempts to autoplace multiple diskful volumes with a storage-only constraint. A bug resulted in the creation of unrelated volumes and a TieBreaker, indicating a meaningless placement.

```bash
r c <rd> --auto-place=2 -l STORAGE
```

--------------------------------

### Blockstor Development Stand Quick Start

Source: https://github.com/cozystack/blockstor/blob/main/AGENTS.md

Commands to bring up, configure, and manage a Blockstor development cluster using the 'make up' command. Includes options for real-disk pools and multiple parallel clusters.

```sh
# Single cluster (default name "blockstor")
make up
make piraeus
make blockstor              # install the controller + satellite DaemonSet
make smoke-blockstor
make down

# Real-disk pools (ZFS + LVM-thin) on extra disks
make pools
STORPOOL=zfs-thin make smoke-blockstor

# Multiple parallel clusters — each gets its own 10.<slot>.0.0/24 CIDR
make up   NAME=alice
make up   NAME=bob
```

--------------------------------

### Example CLI Command for Autoplace with Multiple Pools

Source: https://github.com/cozystack/blockstor/blob/main/docs/cli-parity-audit-2026-05-19-refresh.md

This command attempts to autoplace a resource across multiple storage pools. A bug was found where it only selected one pool due to pre-filtering of the candidate set.

```bash
linstor autoplace --storage-pool=A,B
```

--------------------------------

### Create, List, and Inspect Snapshots

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Demonstrates creating a snapshot, listing snapshots for a specific resource, and listing all cluster-wide snapshots.

```bash
linstor snapshot create test snap1
linstor snapshot list -r test
linstor snapshot list   # cluster-wide
```

--------------------------------

### Simulate Disk I/O Failures with dmsetup

Source: https://github.com/cozystack/blockstor/blob/main/tests/advanced-config-scenarios.md

This example shows how to use dmsetup to create a failing device mapper target, simulating per-sector I/O errors on a block device. This is useful for testing fault tolerance and event handling in storage systems.

```bash
# Wrap the backing LV with a failing dm device:
size=$(blockdev --getsz /dev/blockstor-lvm/some-lv)
# Build a table with a corrupt 1MB region at offset 100MB:
cat <<EOF > /tmp/fail.table
0 204800 linear /dev/blockstor-lvm/some-lv 0
204800 2048 error
206848 $((size-206848)) linear /dev/blockstor-lvm/some-lv 206848
EOF
dmsetup create faildev < /tmp/fail.table

# Reroute the satellite to use the faildev (would require pointing
# blockstor's StoragePool at /dev/mapper/faildev instead of the
# raw LV — design TBD)

# Force a read of the bad region:
dd if=/dev/drbd<N> bs=4096 skip=25000 count=1 of=/dev/null
# Expected: I/O error, DRBD events2 emits `change device disk:Failed`,
# observer auto-detaches per the on-io-error=detach policy

linstor r l -r <rd>
# Expected: this node's replica shows State=Failed or Diskless
# (after auto-detach)
```

--------------------------------

### Tier 2 Integration Test CI Requirements

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-strategy.md

Specifies the necessary setup for Continuous Integration to run Tier 2 integration tests. This includes caching dependencies and installing required packages.

```bash
#setup-envtest action caches `kube-apiserver` + `etcd`
#apt install -y linstor-client python3-linstor in the job
# Build constraint `//go:build integration` so `go test ./...` does not pull integration tests by default
```

--------------------------------

### Get Peer Node ID using Kubernetes-native fields

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-status-cheatsheet.md

Future Kubernetes-native equivalent to get the peer's DRBD node ID. This method reads a denormalized field on the local Resource CRD, simplifying access.

```bash
# Same answer, but as a denormalised field on the local Resource,
# so tests don't need to know the peer's CRD name:
kubectl get resource "${rd}.${node}" -o json \
  | jq -r --arg p "${peer}" '.status.connections[]? | select(.peerNodeName==$p) | .peerDrbdNodeId'
```

--------------------------------

### Get Node

Source: https://github.com/cozystack/blockstor/blob/main/docs/csi-api-surface.md

Retrieves details of a specific node.

```APIDOC
## GET /v1/nodes/{node}

### Description
Retrieves details of a specific node.

### Method
GET

### Endpoint
/v1/nodes/{node}

### Parameters
#### Path Parameters
- **node** (string) - Required - The identifier of the node.
```

--------------------------------

### Mid-level Provisioning with Auto-Placement

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Use `linstor create --auto-place` after manually creating resource and volume definitions to perform only the placement step, allowing the autoplacer to land replicas according to specified counts. This splits the creation process.

```bash
linstor resource-definition create test-auto --resource-group default
linstor volume-definition create test-auto 10G
linstor resource list -r test-auto    # expect empty
linstor create test-auto --auto-place 2
linstor resource list -r test-auto    # expect 3 rows
```

--------------------------------

### Get Peer Disk State using Kubernetes-native fields

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-status-cheatsheet.md

Future Kubernetes-native equivalent to get the peer's disk state. This method reads a denormalized field on the local Resource CRD, providing the local kernel's view of the peer's disk state.

```bash
kubectl get resource "${rd}.${node}" -o json \
  | jq -r --arg p "${peer}" --argjson v "${vol}" \
    '.status.connections[]? \
     | select(.peerNodeName==$p) \
     | .peerVolumes[]? | select(.volumeNumber==$v) | .peerDiskState'
```

--------------------------------

### Get Resource Definition

Source: https://github.com/cozystack/blockstor/blob/main/docs/csi-api-surface.md

Retrieves details of a specific resource definition.

```APIDOC
## GET /v1/resource-definitions/{rd}

### Description
Retrieves details of a specific resource definition.

### Method
GET

### Endpoint
/v1/resource-definitions/{rd}

### Parameters
#### Path Parameters
- **rd** (string) - Required - The identifier of the resource definition.
```

--------------------------------

### Configure Resource Group Placement Constraints

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-ug9-feature-scenarios.md

Example of configuring resource group placement with constraints on replica location. Use `--replicas-on-same` to ensure replicas are on the same site, or `--replicas-on-different` for disaster recovery scenarios.

```bash
--replicas-on-same site --place-count 2
```

```bash
--x-replicas-on-different site 2 --place-count 3
```

--------------------------------

### Get Clone Status

Source: https://github.com/cozystack/blockstor/blob/main/docs/csi-api-surface.md

Retrieves the status of a resource definition cloning operation.

```APIDOC
## GET /v1/resource-definitions/{rd}/clone/{target}

### Description
Retrieves the status of a resource definition cloning operation.

### Method
GET

### Endpoint
/v1/resource-definitions/{rd}/clone/{target}

### Parameters
#### Path Parameters
- **rd** (string) - Required - The identifier of the resource definition.
- **target** (string) - Required - The identifier of the target for the clone operation.
```

--------------------------------

### Manual Provisioning with Explicit Storage Pool

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Use `linstor resource create <node> <resource> -s <pool>` for low-level provisioning, allowing explicit node and storage pool selection. This is useful for debugging autoplacer behavior or specific placement needs.

```bash
linstor resource-definition create test-manual --resource-group default
linstor volume-definition create test-manual 10G
linstor resource create <worker-1> test-manual -s lvm-thin
linstor resource create <worker-2> test-manual -s lvm-thin
linstor resource list -r test-manual
```

--------------------------------

### Get Resource Definition Sync Status

Source: https://github.com/cozystack/blockstor/blob/main/docs/csi-api-surface.md

Retrieves the synchronization status of a resource definition.

```APIDOC
## GET /v1/resource-definitions/{rd}/sync-status

### Description
Retrieves the synchronization status of a resource definition.

### Method
GET

### Endpoint
/v1/resource-definitions/{rd}/sync-status

### Parameters
#### Path Parameters
- **rd** (string) - Required - The identifier of the resource definition.
```

--------------------------------

### Create Storage Pool

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Demonstrates adding a new storage pool on a specific node using LVM_THIN driver.

```bash
linstor sp create <worker-1> new-pool LVM_THIN \
    --pool-name=blockstor-lvm/thin
linstor sp list
```

--------------------------------

### Get Node Storage Pool

Source: https://github.com/cozystack/blockstor/blob/main/docs/csi-api-surface.md

Retrieves details of a specific storage pool on a node.

```APIDOC
## GET /v1/nodes/{node}/storage-pools/{pool}

### Description
Retrieves details of a specific storage pool on a node.

### Method
GET

### Endpoint
/v1/nodes/{node}/storage-pools/{pool}

### Parameters
#### Path Parameters
- **node** (string) - Required - The identifier of the node.
- **pool** (string) - Required - The identifier of the storage pool.
```

--------------------------------

### Configure Topology-Aware Replica Placement

Source: https://github.com/cozystack/blockstor/blob/main/tests/advanced-config-scenarios.md

Sets up Kubernetes node labels for zones and creates a StorageClass with the 'replicasOnDifferent' parameter to ensure replicas are spread across different failure domains. Verifies the placement of replicas.

```bash
kubectl label node worker-1 worker-2 topology.kubernetes.io/zone=a
kubectl label node worker-3 worker-4 topology.kubernetes.io/zone=b
kubectl label node worker-5 worker-6 topology.kubernetes.io/zone=c
```

```bash
# Verify the operator picked up labels into Aux/topology:
linstor node list-properties worker-1
# Expected: Aux/topology/topology.kubernetes.io/zone=a

# StorageClass that spreads:
cat <<EOF | kubectl apply -f -
apiversion: storage.k8s.io/v1
kind: StorageClass
metadata: {name: spread-zones}
provisioner: linstor.csi.linbit.com
parameters:
  replicasOnDifferent: "topology.kubernetes.io/zone"
  placementCount: "3"
  storagePool: stand
EOF

# Provision + verify spread
kubectl apply -f testdata/pvc-spread.yaml
linstor r l -r <pvc-vol-id>
```

--------------------------------

### Get Node Network Interface

Source: https://github.com/cozystack/blockstor/blob/main/docs/csi-api-surface.md

Retrieves details of a specific network interface on a node.

```APIDOC
## GET /v1/nodes/{node}/net-interfaces/{nif}

### Description
Retrieves details of a specific network interface on a node.

### Method
GET

### Endpoint
/v1/nodes/{node}/net-interfaces/{nif}

### Parameters
#### Path Parameters
- **node** (string) - Required - The identifier of the node.
- **nif** (string) - Required - The identifier of the network interface.
```

--------------------------------

### Get LINSTOR-NODE Pod

Source: https://github.com/cozystack/blockstor/blob/main/tests/observability-cheat-sheet-scenarios.md

Retrieves pods labeled with app=linstor-node. Used for upstream LINSTOR node identification.

```bash
kubectl get pod -l app=linstor-node -o wide
```

--------------------------------

### DRBD Resource File Header Comment

Source: https://github.com/cozystack/blockstor/blob/main/docs/design-res-rendering-linstor-parity.md

Example of a header comment block that should be rendered at the top of every .res file.

```text
Header comment block at the top of every .res:
```

--------------------------------

### Mount and Check Resource Usage

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Demonstrates mounting a DRBD device and checking its usage status via `linstor resource list`. This is used to verify auto-promotion to Primary.

```bash
# On worker-1:
mkfs.ext4 /dev/drbd<N> && mount /dev/drbd<N> /mnt/test
# From admin shell:
linstor resource list -r test
```

--------------------------------

### Get Blockstor Satellite Pod

Source: https://github.com/cozystack/blockstor/blob/main/tests/observability-cheat-sheet-scenarios.md

Retrieves pods labeled with app=blockstor-satellite in the blockstor-system namespace. Used for blockstor satellite identification.

```bash
kubectl get pod -n blockstor-system -l app=blockstor-satellite -o wide
```

--------------------------------

### Test Harness Skeleton for E2E Scenarios

Source: https://github.com/cozystack/blockstor/blob/main/tests/scenarios/01-api-contract.md

This script sets up the environment for end-to-end tests, including port-forwarding the Blockstor controller and defining the LINSTOR client command. It ensures the test is independently re-runnable and includes cleanup mechanisms.

```bash
# tests/e2e/linstor-cli-<scenario>.sh
set -euo pipefail
WORK_DIR=${1:?work_dir required}
export KUBECONFIG "$WORK_DIR/kubeconfig"
source "$(dirname "$0")/lib.sh"

kubectl port-forward -n blockstor-system svc/blockstor-controller \
    3370:3370 >/tmp/pf-blockstor.log 2>&1 &
PF_PID=$!
trap 'kill $PF_PID 2>/dev/null || true; cleanup' EXIT
sleep 2

LINSTOR="linstor --controllers=127.0.0.1:3370"
# ... per-scenario commands + asserts ...

```

--------------------------------

### Get the role of a resource

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-status-cheatsheet.md

Retrieves the current role of a Blockstor resource on a specific node. Returns an empty string if the field is not set.

```bash
# $1=rd, $2=node — Status.Role landed in a077afcf2 (Phase 11.5.b P0)
status_role() {
    kubectl get resource "${1}.${2}" -o jsonpath='{.status.role}' 2>/dev/null
}
```

--------------------------------

### Run Unit Tests

Source: https://github.com/cozystack/blockstor/blob/main/tests/scenarios/README.md

Execute unit tests for the pkg and internal directories. These tests typically mock dependencies.

```bash
# Unit tests:
go test ./pkg/... ./internal/...
```

--------------------------------

### List Resources for a Resource Definition

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Shows per-node replicas for a given resource definition, including port, usage, connections, state, and creation timestamp. Verifies the correct number of replicas and their states (UpToDate, TieBreaker).

```bash
linstor resource list -r test1
```

--------------------------------

### DRBD Option Override Regression Example

Source: https://github.com/cozystack/blockstor/blob/main/docs/architecture.md

Illustrates a potential regression where `if *src.X { out.X = src.X }` could silently drop explicit-false overrides.

```go
A regression that did `if *src.X { out.X = src.X }` would silently
   drop explicit-`false` overrides, e.g. an RD that intentionally
   sets `AllowTwoPrimaries=false` would inherit a parent RG's `true`.
```

--------------------------------

### Clone Resource Definition

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Illustrates cloning an existing resource definition to create a derived one with its own volumes and replicas.

```bash
linstor rd clone test test-clone
linstor rd l
```

--------------------------------

### Create Resource and Verify PrefNic Usage

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-ug9-feature-scenarios.md

This sequence of commands creates a new LINSTOR resource and verifies that the resource's configuration file (.res) correctly points to the preferred network interface (PrefNic) for communication.

```bash
linstor rd c prefnic-test
linstor vd c prefnic-test 1G
linstor r c worker-{1,2,3} prefnic-test --storage-pool zfs-thin
sleep 5
ssh worker-1 'cat /var/lib/linstor.d/prefnic-test.res | grep address'
```

--------------------------------

### Get DRBD port for a resource

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-status-cheatsheet.md

Retrieves the DRBD port number for a given resource and node. Returns an empty string if the field is not set.

```bash
# $1=rd, $2=node — int / empty if not yet allocated
status_drbd_port() {
    kubectl get resource "${1}.${2}" -o jsonpath='{.status.drbdPort}' 2>/dev/null
}
```

--------------------------------

### Create and Spawn DRBD Resource with Performance Tuning

Source: https://github.com/cozystack/blockstor/blob/main/tests/advanced-config-scenarios.md

Commands to create a new DRBD resource (RD) and associated volume (VD) after tuning controller properties. This step is crucial to test if the new resource honors the performance settings.

```bash
# Spawn a new RD, check .res honours the props:
linstor rd c perf-test
linstor vd c perf-test 1G
linstor rd ap perf-test --place-count 2
```

--------------------------------

### Get DRBD Replication State (Bypass)

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-status-cheatsheet.md

Checks if the DRBD replication state is 'Established' on a specific node. Useful for verifying successful replication.

```bash
on_node "$node" drbdsetup status "$RD" --verbose 2>/dev/null \
  | grep -E 'replication:Established' | head -1
```

--------------------------------

### Cross-Node Snapshot Shipping (LVM-THIN)

Source: https://github.com/cozystack/blockstor/blob/main/tests/scenarios/06-storage-backends.md

This command sequence demonstrates shipping an LVM-THIN snapshot between nodes using the `thin-send-recv` utility.

```bash
thin-send-recv
```

--------------------------------

### Test Harness Skeleton

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Provides a basic bash script structure for end-to-end Linstor CLI tests, including setup, port-forwarding, and cleanup.

```bash
#!/usr/bin/env bash
# tests/e2e/linstor-cli-<scenario>.sh
set -euo pipefail
WORK_DIR=${1:?work_dir required}
export KUBECONFIG "$WORK_DIR/kubeconfig"
source "$(dirname "$0")/lib.sh"

# Port-forward blockstor-controller in the background, kill on exit.
kubectl port-forward -n blockstor-system svc/blockstor-controller \
    3370:3370 >/tmp/pf-blockstor.log 2>&1 &
PF_PID=$!
trap 'kill $PF_PID 2>/dev/null || true; cleanup' EXIT
sleep 2

# linstor CLI talks to the forwarded port via default config.
LINSTOR="linstor --controllers=127.0.0.1:3370"

# … per-scenario commands + asserts …
```

--------------------------------

### Get DRBD node ID for a resource

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-status-cheatsheet.md

Retrieves the DRBD node ID for a given resource and node. Returns an empty string if the field is not set.

```bash
status_drbd_node_id() {
    kubectl get resource "${1}.${2}" -o jsonpath='{.status.drbdNodeId}' 2>/dev/null
}
```

--------------------------------

### Run All Replay Workflows

Source: https://github.com/cozystack/blockstor/blob/main/CLAUDE.md

Executes all replay workflows defined in the operator-harness. Requires setting the BS_URL environment variable and will exit on the first failure.

```bash
# All replay workflows (CI):
for f in tests/operator-harness/replay/*.yaml; do
    BS_URL=http://127.0.0.1:3370 \
        tests/operator-harness/replay-runner.sh dev-kvaps "$f" || exit 1
done
```

--------------------------------

### Get DRBD minor number for a resource

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-status-cheatsheet.md

Retrieves the DRBD minor number for a given resource and node. Returns an empty string if the field is not set.

```bash
status_drbd_minor() {
    kubectl get resource "${1}.${2}" -o jsonpath='{.status.drbdMinor}' 2>/dev/null
}
```

--------------------------------

### Create Replicated Volume Step-by-Step

Source: https://github.com/cozystack/blockstor/blob/main/docs/usage.md

Manually creates a replicated volume by defining resources and volumes explicitly. This method is useful when not using resource groups.

```sh
linstor resource-definition create myvolume
linstor volume-definition create myvolume 10G
linstor resource create myvolume --auto-place 3 --storage-pool data
```

--------------------------------

### Run Single Scenario

Source: https://github.com/cozystack/blockstor/blob/main/tests/linstor-cli-scenarios.md

Command to execute a specific end-to-end test scenario using the 'make e2e' target.

```bash
make e2e NAME=e2e6 SCENARIO=linstor-cli-spawn
```

--------------------------------

### Get DRBD Role (k8s-native)

Source: https://github.com/cozystack/blockstor/blob/main/docs/test-status-cheatsheet.md

Retrieves the role of a DRBD resource replica using `kubectl`. This field is stamped by an observer parsing `drbdsetup events2`.

```bash
kubectl get resource "${rd}.${node}" -o jsonpath='{.status.role}'
# → "Primary"
```