Frontend Apps

What this screen is for

Frontend Apps deploys and tracks frontend components specifically — the static sites and single-page apps that do not fit the same shape as a service deployment. It gives you a dashboard of what is deployed, what is in flight, and what failed.

Before you start

To do this You need
Open the area the Frontend Deployer access permission (frontendDeployerAccessPermission)

Your components must exist in the catalog — see Settings › Development › Projects.

Opening it

  1. Deploy › Frontend Apps in the sidebar.
Route Screen
/frontend-deployer Dashboard
/frontend-deployer/component/:id One app
/frontend-deployer/settings Settings

There is also a Frontend Deployer tab on catalog entities, for one component.

The interface

📸 Screenshot Placeholder: The Frontend Apps dashboard. Mark: (1) the stat cards, (2) the Filter by System control, (3) the component list, (4) the settings icon.

# Control What it does When you’d use it
1 Stat cards — Components, Deployed, In progress, Failed The state of everything at a glance Opening the page
2 Filter by System Narrows to one system Installs with many frontends
3 Component list Each app and its state Finding one
4 Settings Deployment defaults and retention Setup, and tuning later

Read the stat cards as a set. Failed sitting at a steady non-zero number while Deployed grows is the pattern worth catching — it means failures are being worked around rather than fixed.

Settings

Setting What it controls
Deployment history (days) How long deployment history is kept
Poll interval (seconds) How often status is refreshed
Default pipeline mode The mode new deployments start in
Default deploy template The template used to deploy
Default rollback template The template used to roll back
Shared parameters Parameters applied across deployments

Setting the rollback template at the same time as the deploy template is the part people postpone. You will want it configured before you need it, not during.

Walkthroughs

Deploy a frontend app

  1. Deploy › Frontend Apps.
  2. Find the component, using Filter by System if the list is long.
  3. Deploy it.
  4. Watch the state move through In progress. The page refreshes on the poll interval set in Settings.

Investigate a failure

  1. From the dashboard, open the failed component.
  2. Read its deployment history — how long that is kept is a Settings value, so a very old failure may simply have aged out.

Configure defaults

  1. Deploy › Frontend Apps › Settings.
  2. Set the deploy and rollback templates, the pipeline mode, and any shared parameters.
  3. Set the retention and poll interval to match how you actually work.

Scenario

A frontend deploy failed and nobody noticed for two days.

  1. Deploy › Frontend Apps — the Failed card is non-zero. That card is the whole early-warning system, which is why the dashboard is worth opening even when nothing is reported.
  2. Filter by the owning system to find which component.
  3. Open it and read the deployment history to find when it started failing.
  4. If the history does not go back far enough, check Settings › Deployment history (days). Retention that is shorter than your feedback loop hides the very failures you are trying to investigate.
  5. Fix the deployment, then redeploy.
  6. Confirm the Failed card returns to zero. Treat a persistently non-zero Failed count as a broken window — once it is normal, nobody looks at it again.

When it doesn’t work

Symptom Cause How to check Fix
Frontend Apps is not in the submenu Missing the Frontend Deployer access permission Access Explorer Ask for a role granting it
A component is missing from the list It is not in the catalog, or is filtered out The Filter by System control, then Projects Clear the filter; register the component
Status seems slow to update The poll interval is long Settings › Poll interval Shorten it
Deployment history is shorter than expected Retention is a setting Settings › Deployment history (days) Increase it
Rollback is unavailable No rollback template configured Settings › Default rollback template Configure it before you need it
Failed count never returns to zero Failures are being worked around, not fixed The component list Investigate each; a permanently red card stops being read

Next