cwx/op-packages

0 Fork 0

mirror of https://github.com/kiddin9/op-packages.git synced 2026-10-27 11:52:51 +08:00

Files

T

github-actions[bot] f62992fd80 💋 Sync 2026-10-04 04:16:09

2026-10-04 04:16:09 +08:00

21 KiB

Raw Blame History

Manual UI Testing Plan

0. Authentication

0.1 Navigate to http://<router>/cgi-bin/luci → Redirected to login page
0.2 Submit correct credentials → Redirected to LuCI dashboard
0.3 Navigate to /admin/podman → Overview page loads without re-auth
0.4 Open browser Console tab → No JS errors

1. Overview (`/admin/podman`)

1.1 Load overview → Page loads; all counters visible (0 is fine)
1.2 Podman version string visible → Not empty, not "undefined"
1.3 System status indicator visible → Shows running (socket available)
1.4 Note all current counts → Baseline for comparison after later steps

Prereq: at least one container with label io.containers.autoupdate=registry.

1.5 Click "Check for Updates" → Spinner per container → "Apply Updates" modal lists candidates
1.6 Tick a container with an actually-newer image upstream → Click Update → Modal switches to log view with Update image: 1/N - <name> header, spinner running, Close button disabled
1.7 Log shows Pulling <ref>, then live Trying to pull..., Copying blob ... lines - progress lines with CR overwrite the previous line (no flood of "Copying blob X 50%, 51%, 52%, ...")
1.8 After pull, log shows stage markers: → Stopping container, → Removing container, → Creating container, → Starting container, → Reinstating init script, ✓ Updated to <id-short>
1.9 Multi-container update: each container delimited by ━━ <name> ━━; previous container's log remains visible above current one
1.10 After loop ends, counter shows Done., spinner stops, Close button enabled
1.11 Click Close → modal dismisses → summary alert appears (Containers updated successfully for full success, list of failures otherwise)
1.12 Pull failure (rename the container's image to a non-existent registry before running, or change tag to bogus): log shows ✗ <error>, run continues to next container, summary marks failed
1.13 Mid-flight failure (SSH podman run -d --name <same-name> alpine sleep 3600 between Stop and Create, or simulate by other means): log shows ✗ Create failed: ... for that container, summary marks failed
1.14 Old-image cleanup: after successful update of one container, log shows ━━ Cleaning up old images... ━━ then either ✓ Removed old image <id> or ! Old image <name> could not be removed - still in use by another container
1.15 Modal scroll: when log overflows visible area, content auto-scrolls to bottom on append; scroll up manually → new lines append at bottom but viewport stays where the user put it (sticky-scroll)
1.16 Long pull crossing uhttpd script_timeout (~55s): server-side connection rotates transparently (auto-reconnect in Model._stream); log keeps growing, no spurious error

2. Images (`/admin/podman/images`)

2.1 Load images list → Page loads (may be empty)
2.2 Click "Pull Image" → Form/modal appears with name input
2.3 Submit empty form → Validation error, no request sent
2.4 Pull nonexistent image (e.g. this-does-not-exist:fake) → Error notification shown, list unchanged
2.5 Pull alpine:latest → Progress indicator → success notification → alpine appears in list
2.6 Pull alpine:latest again (already exists) → Graceful result (no crash, may show "already up to date")
2.7 Click inspect on alpine → Modal opens with JSON image data
2.8 Return to overview → Image count incremented by 1

3. Containers (`/admin/podman/containers`)

Requires alpine:latest from step 2.5.

3a. Create

3.1 Load containers list → Page loads (may be empty)
3.2 Click "Create Container" → Form modal opens
3.3 Submit empty form → Validation errors on required fields
3.4 Fill name: test-alpine, image: alpine:latest, no other changes → Submit → Container created; test-alpine visible in list
3.5 Return to overview → Container count incremented

3b. List operations

3.6 Verify test-alpine row shows correct Name, Image, State
3.7 Select test-alpine checkbox → Only that row selected
3.8 Click top "select all" checkbox → All rows selected
3.9 Click top checkbox again → All deselected

3c. Start / Stop / Restart

3.10 If container is stopped: select → Start → State changes to Running; success notification
3.11 Select running container → Stop → State changes to Exited; success notification
3.12 Select stopped container → Restart → State changes to Running

3d. Detail view - Info tab

3.13 Click test-alpine name link → Navigates to container detail page
3.14 Name, image, state displayed → Matches list view
3.15 Edit container name to test-alpine-renamed → Save → Name updated; notification shown
3.16 Change restart policy to always → Save → Policy updated; success notification
3.17 Init script auto-generated → Auto-start indicator shows enabled
3.18 Click "Show Init Script" → Modal with script content; script references test-alpine-renamed
3.19 Toggle init script disabled → Indicator changes
3.20 Toggle init script enabled again → Indicator reverts
3.21 Change restart policy to no → Save → Init script removed/disabled
3.22 Navigate back to list → Name shown as test-alpine-renamed

3e. Detail view - Resources tab

3.23 Click Resources tab → Tab activates; current limits shown
3.24 Set CPU limit to 0.5 → Save → After page reload, Resources tab is active (tab persistence)
3.25 Set Memory limit to 64m → Save → Value persisted
3.26 Enter invalid CPU value (e.g. abc) → Save → Validation error; no save

3f. Detail view - Stats tab

3.27 Start container (if stopped)
3.28 Click Stats tab → CPU%, Memory, Net I/O data appears
3.29 Stop container; click Stats tab → Graceful empty/zero state, no JS crash

3g. Detail view - Logs tab

3.30 Click Logs tab → Log output area visible (may be empty for alpine)
3.31 Change line count to 50 → Apply → Fetches last 50 lines
3.32 Enable live streaming toggle → Toggle activates
3.33 Disable live streaming → Streaming stops, toggle off
3.34 Click Clear → Log area clears

3h. Detail view - Processes tab

3.35 Start container; click Processes tab → Process table appears
3.36 Stop container; click Processes tab → Graceful empty/error state, no JS crash

3i. Network management (on detail page)

3.37 Note current networks attached → Baseline
3.38 Disconnect default network → Network removed from list
3.39 Reconnect the network → Network added back
3.40 Connect a network with valid static IP → Network added with IP shown
3.41 Connect a network with invalid IP → Validation error

3j. Pause / Unpause

3.42 List view: select a running container → click ⏸⏸ Pause → state changes to paused; success notification
3.43 List view: same paused container selected → click ▶ Start → state changes back to running (Start polymorphically unpauses paused containers)
3.44 Detail view: container is running → ▶ Start is highlighted as active, ⏸⏸ Pause is clickable
3.45 Detail view: click Pause → reload → state is paused, ⏸⏸ Pause is now highlighted, ▶ Start is clickable
3.46 Detail view: click Start while paused → reload → state is running, ▶ Start highlighted again
3.47 Detail view: streams (stats / logs / processes) stop before pause fires (no reconnect storm in DevTools Network panel)
3.48 Multi-select on list: mix of running + paused → click ▶ Start → running stays running, paused becomes running (per-container dispatch)

4. Volumes (`/admin/podman/volumes`)

4.1 Load volumes list → Page loads (may be empty)
4.2 Click "Create Volume" → Form appears
4.3 Submit empty form → Name will be generated
4.4 Create volume named test-vol → Appears in list with name, driver, created date
4.5 Click inspect on test-vol → JSON modal shows volume data
4.6 Return to overview → Volume count incremented
4.7 Select test-vol → Delete → Confirmation modal → confirm → removed from list
4.8 Create volume test-vol-prune and leave it → Used in prune test (step 8.2)

5. Networks (`/admin/podman/networks`)

5.1 Load networks list → Default podman network visible
5.2 Note OpenWrt column for default network → Status icon shown
5.3 Click "Create Network" → Form opens
5.4 Submit empty form → Validation error
5.5 Create test_net, subnet 10.220.0.0/24, gateway 10.220.0.1, OpenWrt integration checked → Network created; OpenWrt column shows ✓
5.6 Verify UCI config created → SSH: uci show network.test_net returns values
5.7 Create test_net_b with no subnet, integration unchecked → Network created; OpenWrt column shows -
5.8 If any network shows X (incomplete integration): click it → setup → icon changes to ✓
5.9 Return to overview → Network count incremented
5.10 Select test_net → Delete → Confirmation mentions OpenWrt cleanup → confirm → network removed; UCI config removed
5.11 Select test_net_bare → Delete → Removed cleanly (no UCI to clean up)

6. Secrets (`/admin/podman/secrets`)

6.1 Load secrets list → Page loads (may be empty)
6.2 Click "Create Secret" → Form with name + value fields
6.3 Submit empty form → Validation error
6.4 Create secret name test-secret, value hunter2 → Appears in list; value NOT shown anywhere
6.5 Click inspect on test-secret → Modal shows metadata; SecretData field hidden
6.6 Delete test-secret → Removed from list

7. Pods (`/admin/podman/pods`)

Requires alpine:latest from step 2.5. If a user-defined network exists from section 5, note its name as <usernet>.

7a. List view

7.1 Load pods list → Page loads with empty/populated table; no "Work in progress…" stub
7.2 Toolbar visible → Create / Delete / Reload / ▶ / ■ / ⟳ / ⏸ / ⏵ buttons present
7.3 Click Reload with empty list → No error, indicator appears briefly

7b. Create - minimal

7.4 Click "Create" → Pod form modal opens with default fields populated
7.5 Submit empty form → No validation error (name auto-generates); pod appears with generated name
7.6 Create pod name pod-min, all other defaults → Modal closes, success notification, pod-min row appears
7.7 Verify SSH podman pod inspect pod-min → Pod exists with one infra container
7.8 Containers column shows 1 ▸ for pod-min

7c. Create - full options

7.9 Open Create form → Name pod-full, Hostname pod-full-host, Shared Namespaces cgroup,ipc,net,uts, Infra ON, Network <usernet>, DNS Servers 1.1.1.1 and 9.9.9.9, DNS Search lan, CPU Limit 1.0, CPU Set 0-1, Memory 256m, Labels app=demo\nenv=test → Submit
7.10 SSH podman pod inspect pod-full | jq '.Hostname' → pod-full-host
7.11 SSH podman pod inspect pod-full | jq '.InfraConfig.NetworkOptions' → contains DNS server + search entries
7.12 SSH podman pod inspect pod-full | jq '.Labels' → app: demo, env: test
7.13 Networks column for pod-full → shows <usernet>

7d. Status badges

7.14 New pod → Status badge Created (gray)
7.15 After Start → Badge Running (green)
7.16 After Pause → Badge Paused (orange)
7.17 After Stop → Badge Exited or Stopped (red)

7e. Lifecycle - single select

7.18 Tick pod-min, click ▶ Start → Loading "Starting pod: 1/1", refresh shows Running
7.19 Click ■ Stop → Status Exited/Stopped
7.20 Click ⟳ Restart → Status Running
7.21 Click ⏸ Pause → Status Paused
7.22 Click ⏵ Unpause → Status Running

7f. Lifecycle - multi-select

7.23 Create pod-a and pod-b (defaults), select both → ▶ Start
7.24 Loading modal shows "Starting pod: 1/2" then "2/2"
7.25 Both pods end up Running
7.26 No JS console errors during multi-action

7.27 Click pod Name in list → Navigates to /admin/podman/pod/<id> (no more inspect modal)
7.28 Browser back button → Returns to pods list with state preserved

7.29 SSH podman run -d --pod pod-min --name pod-min-c1 alpine sleep 3600 → container created in pod
7.30 Reload pods list → pod-min Containers column shows 2 ▸
7.31 Click 2 ▸ link → Modal "Containers in pod pod-min" opens with 2 rows (infra + new)
7.32 Each row shows Name, truncated ID, Status badge, Restarts count
7.33 Click container ID link in modal → Navigates to /admin/podman/container/<id> detail view

7i. Pod removal

7.34 Stop pod-min first → Status Exited
7.35 Tick pod-min → Delete → Confirm
7.36 SSH podman pod ls → pod-min gone
7.37 SSH podman ps -a → pod-min-c1 also gone (cascade delete via force)

7j. Container form integration

7.38 Navigate Podman → Containers → Create → Form shows new Pod ListValue (after Image) with "(none)" + existing pod names
7.39 Select Pod = pod-full → Network, Hostname, Port Mappings, Expose Ports fields hide
7.40 Select Pod = "(none)" → Hidden fields reappear
7.41 Create container with Pod pod-full, Image alpine:latest, Command sleep 3600, Start ON → Success
7.42 SSH podman ps --filter pod=pod-full → New container listed
7.43 Pods list pod-full Containers column → count incremented

7k. Edge cases

7.44 Create pod with name Bad Name! → Validation error (uciname constraint)
7.45 Toggle "Create Infra Container" OFF → Network/DNS/Infra Image fields hide
7.46 Submit pod with Infra OFF → Pod created without infra container; SSH podman pod inspect confirms
7.47 Try Stop on already-stopped pod → No JS error (graceful handling of API 304/409)
7.48 Try Pause on stopped pod → Error notification, no JS crash

7l. Regression - non-pod container create

7.49 Create container with Pod = "(none)", port mapping 8080:80, network <usernet>, hostname freebee → Success
7.50 SSH inspect → port mapping, network and hostname present (confirms if (!data.pod) guards don't break the non-pod path)

7m. Cleanup

7.51 SSH podman pod rm -f pod-full pod-a pod-b → All test pods removed
7.52 SSH podman ps -a → No leftover test containers
7.53 Reload pods list → Empty/clean state

7n. Detail page - load & header (`/admin/podman/pod/<id>`)

7.54 Direct URL /admin/podman/pod/<podId> → Detail page loads, no console errors
7.55 Header shows pod name and back-arrow button → back button returns to pods list
7.56 Lifecycle buttons visible: ▶ Start, ■ Stop, ⟳ Restart, ⏸ Pause, ⏵ Unpause, 🗑 Delete
7.57 "Active" highlight reflects current state: Running pod → ▶ highlighted, Paused pod → ⏸ highlighted, Exited pod → ■ highlighted (semantic: button describes current state, not "available action")
7.58 Tabs visible: Info, Stats, Processes, Inspect

7o. Detail page - Info tab

7.59 Basic Information section: Name, ID (full 64-char), Status badge, Created date populated
7.60 Configuration section: Hostname, Cgroup parent, Exit policy, Restart policy, Shared namespaces (comma-joined), Infra container (clickable link to /admin/podman/container/<infraId>)
7.61 Resources section: CPU period (µs), CPU quota (µs or "Unlimited"), CPU shares, CPU set, Memory limit (formatted bytes or "Unlimited"). Description note "Resources are set at pod creation and cannot be edited here." present
7.62 Network section: Networks, Static IPv4, Static MAC, DNS servers, DNS search, DNS options, Extra hosts, Host network (Yes/No)
7.63 Containers section: table with Name, truncated ID (clickable → container detail), Status badge - one row per container including infra
7.64 Labels section: only rendered if pod has labels; otherwise omitted
7.65 All empty fields show - placeholder

7p. Detail page - Stats tab (pod running)

7.66 Switch to Stats tab → Table with headers: Container, CID, CPU %, Memory, Memory %, Net I/O, Block I/O, PIDs
7.67 Rows update live (humanized strings from PodStatsReport: e.g. "75.5%", "12mb / 24mb")
7.68 Container name column resolves from pod's containers map (not the pod-name value the API returns)
7.69 Switch to another tab → Stream stops (verify in DevTools Network: stream connection closes)
7.70 Switch back to Stats → Stream restarts
7.71 Pod is stopped → Stats tab shows "Pod is not running" warning, no stream attempt

7q. Detail page - Processes tab (pod running)

7.72 Switch to Processes tab → Table with columns Podman's pod top returns (USER, PID, ..., CMD) plus a CONTAINER column (native to pod top)
7.73 Rows update live as processes change
7.74 Switch away → Stream stops
7.75 Pod is stopped → Shows "Pod is not running" warning

7r. Detail page - Inspect tab

7.76 Switch to Inspect tab → Renders full raw JSON of podman pod inspect
7.77 Same content as old list-view inspect modal (sanity check no fields are lost)

7s. Detail page - lifecycle actions

7.78 Running pod → click ⏸ Pause → confirmation/loading → reload → status Paused, ⏸ now highlighted as active
7.79 Paused pod → click ⏵ Unpause → reload → status Running
7.80 Running pod → click ■ Stop → reload → status Exited, ■ highlighted
7.81 Stopped pod → click ▶ Start → reload → status Running
7.82 Click ⟳ Restart → reload → status briefly cycles to Running
7.83 Click 🗑 Delete → confirm dialog → confirm → redirects to /admin/podman/pods, pod gone
7.84 Before Stop/Pause/Restart, verify in DevTools Network: stats/processes streams close BEFORE the action fires (no reconnect storm)

7t. Detail page - edge cases

7.85 Pod with no containers (rare - infra-less or just created): Containers section shows "No containers in this pod"
7.86 Pod with no labels: Labels section is absent (not just empty)
7.87 Pod with RestartPolicy unset: Configuration shows -
7.88 Pod with memory_limit: 0 or unset: Resources shows "Unlimited"
7.89 Pod with cpu_quota: -1: Resources shows "Unlimited"
7.90 Invalid pod ID in URL → Redirects to pods list (no crash)
7.91 Refresh page while stats stream active → No duplicate streams (Network panel shows single connection)

8. Modals

8.1 Open prune modal → Options shown: containers, images, volumes, networks
8.2 Select only "Volumes" → Prune → Only unused volumes pruned (e.g. test-vol-prune from step 4.8); containers/images untouched
8.3 Select "Containers" → Prune → Only stopped/exited containers removed

8.4 Open auto-update modal → Container list shown with current digest info
8.5 Click "Check for Updates" → Digest comparison runs; up-to-date containers marked
8.6 If outdated container detected → Update → Pulls new image, recreates container

9. Error Handling

9.1 Stop Podman socket (/etc/init.d/podman stop), load any list view → Clear RPC error message; no JS crash, no blank page
9.2 Navigate to /admin/podman/container/nonexistent-id → Graceful error, not a blank page
9.3 Start Podman again (/etc/init.d/podman start); reload → Normal operation resumes
9.4 Trigger two operations quickly in succession → No race-condition crash or duplicate requests

10. Cleanup Verification

10.1 Delete any remaining test containers (e.g. test-alpine-renamed)
10.2 Delete alpine image
10.3 Containers list → Empty (or back to pre-test state)
10.4 Images list → Empty (or back to pre-test state)
10.5 Overview counters → Match baseline from step 1.4

21 KiB Raw Blame History