# Phase 09-04 Deployment and Test Plan

**Generated:** 2026-02-04
**Purpose:** Verification testing for batch operations implementation

## Deployment Steps

### 1. Import Updated Workflow

1. Open your n8n instance (typically at http://your-server:5678)
2. Navigate to Workflows
3. Select "Docker Manager Bot" workflow
4. Click the three-dot menu → Export
5. Save current version as backup: `n8n-workflow-backup-20260204.json`
6. Return to Workflows → Import from File
7. Select the updated `n8n-workflow.json` from this repository
8. Confirm credential mapping (should use existing "Telegram API" credential)
9. Save and activate the workflow

### 2. Verify Workflow Health

Before testing, confirm:
- ✅ Workflow is "Active" (toggle in top-right)
- ✅ No error indicators on nodes
- ✅ Telegram Trigger shows connected
- ✅ Your user ID is still configured in IF User/Callback Authenticated nodes

## Test Plan

### Test Suite A: Batch Text Commands

#### Test A1: Multi-Container Update
**Objective:** Verify batch update with space-separated names

**Steps:**
1. Send to bot: `update plex sonarr` (use 2-3 of your actual container names)
2. Observe behavior

**Expected Results:**
- ✅ Message shows "Updating N containers..."
- ✅ Progress updates appear for each container individually
- ✅ Shows "Pulling image..." → "Stopping..." → "Starting..." per container
- ✅ Final summary message appears with success count
- ✅ If any container fails, shows failure with reason but continues batch
- ✅ Summary emphasizes failures (if any) over successes

**Notes:**
- Record actual execution time for performance assessment
- Screenshot final summary for documentation

---

#### Test A2: Multi-Container Start
**Objective:** Verify batch start executes immediately (no confirmation)

**Preparation:**
1. Manually stop 2-3 containers via existing single commands: `stop plex`, `stop sonarr`
2. Confirm containers are stopped via `status`

**Steps:**
1. Send to bot: `start plex sonarr` (use your stopped container names)
2. Observe behavior

**Expected Results:**
- ✅ No confirmation prompt (starts immediately)
- ✅ Progress shows for each container
- ✅ Summary shows successful starts
- ✅ Verify containers are running via `status`

---

#### Test A3: Multi-Container Stop with Confirmation
**Objective:** Verify batch stop requires confirmation (safety measure)

**Steps:**
1. Send to bot: `stop plex sonarr` (use 2+ running containers)
2. Wait for confirmation prompt
3. Tap "Confirm" button
4. Observe execution

**Expected Results:**
- ✅ Confirmation message appears: "Stop 2 containers?"
- ✅ Lists container names in confirmation
- ✅ Has "Confirm" and "Cancel" buttons
- ✅ After confirm: batch execution proceeds with progress
- ✅ Summary shows containers stopped

**Follow-up Test:**
1. Repeat but tap "Cancel" button
2. Expected: Confirmation deleted, no action taken

**Follow-up Test 2:**
1. Repeat but don't respond for 30+ seconds
2. Expected: Confirmation expires with message

---

#### Test A4: Fuzzy Matching with Exact Match Priority
**Objective:** Verify exact match takes priority over partial matches

**Scenario 1: Exact match exists**
1. Send: `update plex` (when both "plex" and "jellyplex" exist)
2. Expected: Only "plex" container updates (no disambiguation)

**Scenario 2: Only partial matches**
1. Send: `update jelly` (matches "jellyplex" but not exact)
2. Expected: If only one match, proceeds; if multiple, shows disambiguation

**Note:** This test depends on your actual container names. Adjust to match your server.

---

#### Test A5: Disambiguation for Ambiguous Names
**Objective:** Verify disambiguation prompt appears when multiple containers match

**Steps:**
1. Send command with ambiguous partial match (e.g., `update lin` if you have multiple "lin*" containers)
2. Wait for disambiguation prompt
3. Select intended container

**Expected Results:**
- ✅ Shows "Multiple containers match: lin"
- ✅ Lists matching containers with buttons
- ✅ Selecting one proceeds with single-container action
- ✅ Batch not triggered (user clarified intent)

**Note:** If no ambiguous names on your server, document "Cannot test - no ambiguous container names"

---

### Test Suite B: Update All Command

#### Test B1: Update All with Available Updates
**Objective:** Verify "update all" targets only :latest containers with updates

**Steps:**
1. Send: `update all` (or `updateall`)
2. Observe confirmation prompt
3. Note which containers are listed
4. Tap "Confirm"
5. Observe execution

**Expected Results:**
- ✅ Command recognized and routed
- ✅ Confirmation shows: "Update N containers?"
- ✅ Lists containers (max 10 displayed in message)
- ✅ Only includes containers using :latest tag
- ✅ 30-second timeout on confirmation
- ✅ After confirm: batch execution with progress per container
- ✅ Summary shows results

**Verification:**
- Check that only :latest containers were updated
- Containers with specific tags (e.g., `:1.2.3`) should not appear in list

---

#### Test B2: Update All When No Updates Available
**Objective:** Verify appropriate message when all containers are current

**Preparation:**
1. Update all containers manually first OR wait until all are current

**Steps:**
1. Send: `update all`
2. Observe response

**Expected Results:**
- ✅ Shows: "All containers are up to date!" (or similar message)
- ✅ No confirmation prompt
- ✅ No batch execution attempted

---

#### Test B3: Update All Cancel
**Objective:** Verify cancel works

**Steps:**
1. Send: `update all`
2. Wait for confirmation
3. Tap "Cancel"

**Expected Results:**
- ✅ Confirmation message deleted
- ✅ Shows cancellation feedback
- ✅ No containers updated

---

#### Test B4: Update All Timeout
**Objective:** Verify expiration behavior

**Steps:**
1. Send: `update all`
2. Wait for confirmation
3. Don't respond for 30+ seconds
4. Try tapping button after expiry

**Expected Results:**
- ✅ After 30s: Shows expiry message
- ✅ Confirmation becomes inactive
- ✅ Tapping expired button shows alert

---

### Test Suite C: Inline Keyboard Multi-Select

#### Test C1: Enter Multi-Select Mode
**Objective:** Verify multi-select keyboard appears

**Steps:**
1. Send: `/status`
2. Locate "Select Multiple" button (may need to be added in future plan)
3. OR send callback manually: Use bot command that triggers `batch:mode`

**Note:** If no entry point exists yet, test by:
- Temporarily adding "Select Multiple" button to status keyboard
- OR testing via n8n "Execute Node" with callback_query data: `batch:mode`

**Expected Results:**
- ✅ Keyboard shows container list with state icons
- ✅ Running containers: 🟢
- ✅ Stopped containers: ⚪
- ✅ Each button shows container name
- ✅ No checkmarks initially
- ✅ Bottom row has "Cancel" button

---

#### Test C2: Toggle Selection
**Objective:** Verify checkmarks toggle on/off

**Steps:**
1. Enter multi-select mode (from C1)
2. Tap a container button
3. Observe keyboard update
4. Tap same container again
5. Tap different container

**Expected Results:**
- ✅ First tap: Checkmark (✓) appears before container name
- ✅ Second tap: Checkmark disappears
- ✅ Multiple containers can have checkmarks
- ✅ Action buttons appear when any container selected
- ✅ "Clear Selection" button appears with selection

---

#### Test C3: Execute Batch Update from Multi-Select
**Objective:** Verify batch execution from inline keyboard

**Steps:**
1. Enter multi-select mode
2. Select 2-3 containers (tap to add checkmarks)
3. Tap "Update Selected (N)" button
4. Observe execution

**Expected Results:**
- ✅ Immediate execution (no confirmation for update)
- ✅ Progress shows for each selected container
- ✅ Summary shows results
- ✅ Selection message deleted after execution starts

---

#### Test C4: Execute Batch Stop with Confirmation
**Objective:** Verify stop requires confirmation from multi-select

**Steps:**
1. Enter multi-select mode
2. Select 2+ running containers
3. Tap "Stop Selected (N)" button
4. Wait for confirmation
5. Tap "Confirm"

**Expected Results:**
- ✅ Confirmation prompt appears (doesn't execute immediately)
- ✅ Lists selected containers
- ✅ After confirm: batch stop executes
- ✅ Summary shows stopped containers

---

#### Test C5: Selection Limit Enforcement
**Objective:** Verify callback size limit prevents overflow

**Steps:**
1. Enter multi-select mode
2. Select containers one by one
3. Attempt to select 9+ containers

**Expected Results:**
- ✅ Selection works smoothly for first ~8 containers
- ✅ At limit: Alert appears "Selection limit reached"
- ✅ Cannot select additional containers
- ✅ Can deselect and select different containers
- ✅ Guidance shown (e.g., "Use 'update all' for larger batches")

**Note:** Exact limit depends on container name lengths. Shorter names = more selections possible.

---

#### Test C6: Clear Selection
**Objective:** Verify clear button resets selection

**Steps:**
1. Select 3+ containers
2. Tap "Clear Selection" button

**Expected Results:**
- ✅ All checkmarks removed
- ✅ Action buttons disappear
- ✅ Only "Cancel" button remains
- ✅ Can start new selection

---

#### Test C7: Cancel Multi-Select
**Objective:** Verify cancel exits cleanly

**Steps:**
1. Enter multi-select mode (with or without selection)
2. Tap "Cancel" button

**Expected Results:**
- ✅ Selection message deleted
- ✅ Returns to previous state
- ✅ No actions executed

---

### Test Suite D: Error Handling

#### Test D1: Failure Isolation
**Objective:** Verify one failure doesn't abort batch

**Steps:**
1. Create a batch with intentional failure (e.g., non-existent container mixed with real ones)
2. Example: `update plex nonexistent sonarr`
3. Observe execution

**Expected Results:**
- ✅ First container processes
- ✅ Failed container shows error message with reason
- ✅ Remaining containers still process (batch continues)
- ✅ Summary shows: "2 succeeded, 1 failed"
- ✅ Failure details prominent in summary

---

#### Test D2: Warning vs Error Classification
**Objective:** Verify non-critical warnings don't show as errors

**Setup:**
1. Stop a container: `stop plex`
2. Try stopping again: `stop plex`

**Expected Results:**
- ✅ Shows as warning, not error
- ✅ Message: "Already stopped" or similar
- ✅ Summary distinguishes warnings from errors

**Similar tests:**
- Update container with no update available → Warning, not error
- Start already-running container → Warning

---

### Test Suite E: Regression Tests

#### Test E1: Single-Container Commands Still Work
**Objective:** Verify no regression in existing functionality

**Commands to test:**
1. `status` → Shows container list keyboard
2. `start plex` → Starts single container
3. `stop plex` → Shows confirmation, then stops
4. `restart plex` → Restarts with progress
5. `update plex` → Updates single container
6. `logs plex` → Shows last 50 lines
7. `logs plex 100` → Shows last 100 lines

**Expected Results:**
- ✅ All single-container commands work exactly as before
- ✅ No batch behavior triggered for single containers
- ✅ Confirmation behavior unchanged (stop requires confirm, others don't)

---

#### Test E2: Inline Keyboard Actions Still Work
**Objective:** Verify phase 8 keyboard functionality intact

**Steps:**
1. Send: `status`
2. Tap a container name button
3. Tap an action button (e.g., "▶️ Start", "Update 🔄")
4. Complete action

**Expected Results:**
- ✅ Container detail view appears
- ✅ Action buttons work
- ✅ Actions execute correctly
- ✅ No interference from batch infrastructure

---

#### Test E3: Pagination Still Works
**Objective:** Verify container list pagination for many containers

**Steps:**
1. Send: `status` (if you have 10+ containers)
2. Navigate with Previous/Next buttons

**Expected Results:**
- ✅ Pagination works correctly
- ✅ Page numbers accurate
- ✅ No batch selection interference

**Note:** If < 10 containers, document "Cannot test - insufficient containers"

---

## Success Criteria Verification

After completing all tests, verify these criteria are met:

- [ ] **BAT-01:** User can update multiple containers in one command
  - Tests: A1, C3

- [ ] **BAT-02:** Batch updates execute sequentially with per-container feedback
  - Tests: A1, A2, C3

- [ ] **BAT-03:** "Update all" updates only containers with updates available
  - Tests: B1, B2

- [ ] **BAT-04:** "Update all" requires confirmation
  - Tests: B1, B3, B4

- [ ] **BAT-05:** One failure doesn't abort remaining batch
  - Tests: D1

- [ ] **BAT-06:** Final summary shows success/failure count
  - Tests: A1, C3, D1

- [ ] **Inline keyboard batch selection works**
  - Tests: C1-C7

- [ ] **No regression in existing commands**
  - Tests: E1, E2, E3

## Test Execution Log

**Date:** ___________
**Tester:** ___________
**n8n Version:** ___________
**Workflow Import Time:** ___________

### Results Summary

| Test | Status | Notes |
|------|--------|-------|
| A1: Multi-container update | ⬜ Pass / ⬜ Fail | |
| A2: Multi-container start | ⬜ Pass / ⬜ Fail | |
| A3: Multi-container stop | ⬜ Pass / ⬜ Fail | |
| A4: Fuzzy matching | ⬜ Pass / ⬜ Fail / ⬜ N/A | |
| A5: Disambiguation | ⬜ Pass / ⬜ Fail / ⬜ N/A | |
| B1: Update all with updates | ⬜ Pass / ⬜ Fail | |
| B2: Update all (none available) | ⬜ Pass / ⬜ Fail / ⬜ N/A | |
| B3: Update all cancel | ⬜ Pass / ⬜ Fail | |
| B4: Update all timeout | ⬜ Pass / ⬜ Fail | |
| C1: Enter multi-select | ⬜ Pass / ⬜ Fail | |
| C2: Toggle selection | ⬜ Pass / ⬜ Fail | |
| C3: Batch update from multi-select | ⬜ Pass / ⬜ Fail | |
| C4: Batch stop with confirm | ⬜ Pass / ⬜ Fail | |
| C5: Selection limit | ⬜ Pass / ⬜ Fail | |
| C6: Clear selection | ⬜ Pass / ⬜ Fail | |
| C7: Cancel multi-select | ⬜ Pass / ⬜ Fail | |
| D1: Failure isolation | ⬜ Pass / ⬜ Fail | |
| D2: Warning vs error | ⬜ Pass / ⬜ Fail | |
| E1: Single commands regression | ⬜ Pass / ⬜ Fail | |
| E2: Inline keyboard regression | ⬜ Pass / ⬜ Fail | |
| E3: Pagination regression | ⬜ Pass / ⬜ Fail / ⬜ N/A | |

### Issues Found

| Issue # | Test | Description | Severity | Status |
|---------|------|-------------|----------|--------|
| | | | | |

### Additional Observations

_Record any unexpected behavior, performance notes, UX feedback, etc._

---

**Overall Assessment:** ⬜ Ready for Production / ⬜ Issues Need Resolution

**Notes for Phase 10:**
_Record any polish/improvement ideas discovered during testing_