Skip to content

docs(model-serving): update for deployment-only create modal, revision UX, and detail page polish#7443

Open
yomybaby wants to merge 2 commits into
05-15-docs_rename_wsproxy_to_app_proxy_across_the_manualfrom
05-15-docs_model-serving_update_for_deployment-only_create_modal_revision_ux_and_detail_page_polish
Open

docs(model-serving): update for deployment-only create modal, revision UX, and detail page polish#7443
yomybaby wants to merge 2 commits into
05-15-docs_rename_wsproxy_to_app_proxy_across_the_manualfrom
05-15-docs_model-serving_update_for_deployment-only_create_modal_revision_ux_and_detail_page_polish

Conversation

@yomybaby
Copy link
Copy Markdown
Member

@yomybaby yomybaby commented May 16, 2026

resolves #NNN (FR-MMM)

Rewrites the Model Serving documentation across all four locales (EN, JA, KO, TH) to reflect the new two-step deployment creation flow and the revised Endpoint Detail Page structure.

Creation flow changes:

  • Replaces the single "Service Launcher" section with a two-step flow: Create Deployment (deployment-level metadata only) followed by Add Revision (runtime configuration snapshot).
  • Documents the new Create Deployment modal fields: deployment name, Open To Public toggle, and resource group selection.
  • Documents the new Add Revision modal fields: runtime variant, environment/version, model storage, start command (Custom only), environment variables, resource preset, and auto-activate toggle.
  • Notes that the revision name field has been removed; revisions are now identified by auto-assigned revision numbers.
  • Documents Preset Mode for revision creation, including preset selector behavior, field pre-filling, and fallback to manual mode when no compatible presets exist.

Endpoint Detail Page changes:

  • Adds a new Deployment Alerts section documenting all contextual banners: Healthy/ready, private deployment token warning, no current revision, preparing, and not-in-project states.
  • Updates the Service Info card to reflect the new Visibility (Public/Private) tag, Resource Group as deployment-level metadata, and Shared Memory (SHM) in the resources field.
  • Documents the More Menu on the Service Info card and clarifies the filled (DeleteFilled) vs. outlined (DeleteOutlined) trash icon convention for permanent delete vs. soft delete.
  • Adds a Replicas section documenting the new Running/Terminated radio filter and the three independent status fields per replica (Lifecycle, Health, Traffic).
  • Adds a Revisions Tab section documenting revision number ordering, column definitions, filtering/sorting, and the Apply action with its rollout behavior.
  • Updates the service termination instructions to use the More menu → Delete Deployment flow with typed-confirmation modal instead of the Controls column Delete button.

Adds new screenshots for the Create Deployment modal, Add Revision modal, endpoint detail visibility tag, endpoint detail private alert, endpoint detail More menu, and revision history tab across all locales.

Checklist:

  • Documentation
  • Minium required manager version
  • Specific setting for review (eg., KB link, endpoint or how to setup)
  • Minimum requirements to check during review
  • Test case(s) to demonstrate the difference of before/after

@github-actions github-actions Bot added the size:XL 500~ LoC label May 16, 2026
Copy link
Copy Markdown
Member Author

yomybaby commented May 16, 2026

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more


How to use the Graphite Merge Queue

Add either label to this PR to merge it via the merge queue:

  • flow:merge-queue - adds this PR to the back of the merge queue
  • flow:hotfix - for urgent changes, fast-track this PR to the front of the merge queue

You must have a Graphite account in order to use the merge queue. Sign up using this link.

An organization admin has required the Graphite Merge Queue in this repository.

Please do not merge from GitHub as this will restart CI on PRs being processed by the merge queue.

This stack of pull requests is managed by Graphite. Learn more about stacking.

Copy link
Copy Markdown
Contributor

Copilot AI left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

Updates the Model Serving section of the user manual (EN, KO, JA, TH) to reflect the new two-step deployment + revision workflow, the redesigned Endpoint Detail Page (deployment alerts, Visibility/SHM, More menu, Replicas filter, Revisions tab), and the unified "Apply" terminology. Also refreshes the termination instructions to use the new typed-confirmation Delete Deployment flow.

Changes:

  • Replaces the single Service Launcher description with separate Create Deployment, Add Revision, and Preset-mode sections, and keeps the existing field reference as a follow-up subsection.
  • Expands the Endpoint Detail Page with new alert variants, Visibility/Resource Group/SHM fields, More menu/delete-icon conventions, a Replicas tab, and a Revisions tab with an Apply action.
  • Rewrites the "Terminating the Service" guidance to point at the More menu's Delete Deployment flow.

Reviewed changes

Copilot reviewed 4 out of 28 changed files in this pull request and generated 4 comments.

File Description
packages/backend.ai-webui-docs/src/en/model_serving/model_serving.md English source updated for the new deployment/revision flow and Endpoint Detail Page sections.
packages/backend.ai-webui-docs/src/ko/model_serving/model_serving.md Korean translation kept structurally in sync with the English changes.
packages/backend.ai-webui-docs/src/ja/model_serving/model_serving.md Japanese translation kept structurally in sync with the English changes.
packages/backend.ai-webui-docs/src/th/model_serving/model_serving.md Thai translation kept structurally in sync with the English changes.
Comments suppressed due to low confidence (2)

packages/backend.ai-webui-docs/src/en/model_serving/model_serving.md:565

  • endpoint_detail_shm.png does not exist in the repository, and the adjacent TODO comment confirms the screenshot has not been captured yet. The same broken image reference is added to the ko, ja, and th translations. Capture the screenshot or remove the reference before merging to avoid a broken image in published docs.
![](../images/endpoint_detail_shm.png)
<!-- TODO: Capture screenshot — SHM displayed in endpoint detail config -->

packages/backend.ai-webui-docs/src/en/model_serving/model_serving.md:585

  • replica_status_filter.png does not exist anywhere under src/*/images/, and the adjacent TODO confirms it has not been captured. The same broken reference is added to all four language versions. Capture the screenshot or remove the image (and TODO) before merging.
![](../images/replica_status_filter.png)
<!-- TODO: Capture screenshot — Running/Terminated radio filter -->

Comment thread packages/backend.ai-webui-docs/src/en/model_serving/model_serving.md Outdated
Comment thread packages/backend.ai-webui-docs/src/en/model_serving/model_serving.md Outdated
Comment thread packages/backend.ai-webui-docs/src/en/model_serving/model_serving.md Outdated
@agatha197 agatha197 force-pushed the 05-15-docs_rename_wsproxy_to_app_proxy_across_the_manual branch from a6f0429 to deb757e Compare May 18, 2026 03:54
@agatha197 agatha197 force-pushed the 05-15-docs_model-serving_update_for_deployment-only_create_modal_revision_ux_and_detail_page_polish branch 2 times, most recently from 1b25189 to accb898 Compare May 18, 2026 05:20
@agatha197 agatha197 force-pushed the 05-15-docs_rename_wsproxy_to_app_proxy_across_the_manual branch from deb757e to b1307b7 Compare May 18, 2026 05:20
yomybaby added 2 commits May 18, 2026 06:24
…n UX, and detail page polish

Part of the recent main-branch docs catch-up plan (Work Item 1 + folded Work Item 9).

Covers numerous user-visible Model Serving changes merged to main between Nov 2025 and May 2026:

- Creating a Model Service: new deployment-only create modal flow (FR-2822) with resource-group auto-select; redesigned Add Revision modal exposing start command, env vars, runtime variant, resource group, preset, and auto-activate (FR-2826/2835/2836/2886/2888/2889/2891); revision name field removed.
- New H3: Preset Mode for Revision Creation (FR-2862/2863).
- Endpoint Detail Page: new deployment alerts — Deployment Ready with chat shortcut (FR-2830), Private Deployment with access-token shortcut (FR-2838), NoCurrentRevision warning (FR-2843).
- Deployment ID display + Visibility row with Public/Private BooleanTag (FR-2833/2834).
- More menu next to Edit, with DeleteFilled vs DeleteOutlined icon convention (FR-2846/2848).
- Replicas: Running/Terminated radio filter replaces enum status filter, new ModelReplica status fields (FR-2891/2904).
- Shared Memory (SHM) display in endpoint detail config (FR-2837).
- Revisions tab: revision number column, expanded filter/sort, unified "Apply" terminology (FR-2858/2902).
- Cross-reference to forthcoming Deployment Presets page (Work Item 11, lands in PR C).

Updated in all 4 languages (en/ko/ja/th). Screenshots are flagged with TODO markers for separate capture.
Captures 6 of 9 planned screenshots × 4 languages (24 PNGs) for the Model Serving overhaul:
- model_serving_create_modal.png
- model_serving_add_revision_modal.png
- endpoint_detail_private_alert.png
- endpoint_detail_visibility.png
- endpoint_detail_more_menu.png
- revision_history_tab.png

Skipped (TODO markers retained — require specific deployment state not reachable on the test backend in this session):
- endpoint_detail_ready_alert.png (needs a successfully deployed Ready revision)
- endpoint_detail_shm.png (needs revision with custom SHM allocation)
- replica_status_filter.png (needs FR-2904 build visible on backend; visible on local dev but the test backend snapshot lagged behind)

Captured at 2× zoom via Portless dev (https://docs-capture.localhost:1356) against isla-sorna test backend, per-locale via window.switchLanguage().
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

size:XL 500~ LoC

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants