Skip to content

docker-image.yml: manual trigger + decouple client/server deploy hooks - #169

Merged
buzz-code merged 1 commit into
mainfrom
claude/timestamp-utc-offset-prod-8tc72k
Sep 4, 2026
Merged

buzz-code merged 1 commit into
mainfrom
claude/timestamp-utc-offset-prod-8tc72k

Conversation

@buzz-code

Copy link
Copy Markdown
Owner

Summary

Investigated why production wasn't picking up the timezone: 'Z' fix (server/shared config, merged in #166/#167).

Found: docker-image.yml has fired and built+pushed images successfully on every push to main (the earlier claim of "no runs since March" was wrong). The actual break was the deploy-webhook steps (Update client service / Update server service) returning HTTP 500 "Error getting registry": "object not found inside the database" since 2026-09-01 — a Portainer-side registry reference gone stale. That's now fixed on the Portainer side (stack re-saved).

While diagnosing, found a compounding bug in this workflow: a failure in Update client service implicitly skipped Update server service too (GitHub Actions ANDs a step's if: with success() by default), even on server-only commits.

Changes

  • Add workflow_dispatch: {} so a deploy can be re-fired manually without an empty commit.
  • Add !cancelled() to both webhook steps' if: conditions so a failure in one no longer blocks the other.

Test plan

  • Confirmed via Actions run logs that build+push always succeeded; only the webhook steps failed.
  • After merge, push a trivial server-side change to confirmed the full pipeline (build → push → webhook → Portainer redeploy) now goes green end-to-end.

🤖 Generated with Claude Code

https://claude.ai/code/session_01Grpcg4VKLrJw9MkLWUMMBS


Generated by Claude Code

Add workflow_dispatch so a deploy can be re-fired manually without an
empty commit.

Add !cancelled() to the client/server webhook steps' if-conditions.
Previously a failure in "Update client service" implicitly blocked
"Update server service" from running at all (GitHub Actions ANDs a
step's own if-condition with success() unless overridden), even on
commits that only touched server/**. That's why several recent
pushes never attempted the server deploy hook.

Root cause of the actual webhook failures (both CLIENT_WEBHOOK and
SERVER_WEBHOOK returning HTTP 500 "Error getting registry: object
not found inside the database" since 2026-09-01) is on the deploy
target side, not in this workflow - image build+push has succeeded
on every push.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Grpcg4VKLrJw9MkLWUMMBS
Copilot AI lite review requested due to automatic review settings September 4, 2026 07:20
@buzz-code
buzz-code merged commit 42b3310 into main Sep 4, 2026
1 check passed

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

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

🟡 Changes recommended

Manual workflow_dispatch runs likely skip webhook steps due to paths-filter outputs staying false, so stated “manual redeploy” goal not met without condition update.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Pull request overview

Adjust GitHub Actions deploy workflow so deploy can be re-run manually and client/server Portainer webhooks no longer block each other on step failure.

Changes:

  • Add workflow_dispatch trigger to allow manual workflow runs.
  • Update webhook step if: conditions to include !cancelled() so step can run even after earlier step failure.
File summaries
File Description
.github/workflows/docker-image.yml Add manual trigger and tweak webhook step conditions to decouple client/server deploy hooks
Review details

Suppressed comments (1)

.github/workflows/docker-image.yml:65

  • workflow_dispatch run likely sets steps.changes.outputs.server to 'false' (paths-filter compares ref to base), so manual run will skip webhook and not redeploy. Include workflow_dispatch in condition so webhook can be re-fired manually.
      - name: 'Update server service'
        if: steps.changes.outputs.server == 'true' && !cancelled()
        uses: fjogeleit/http-request-action@v1
  • Files reviewed: 1/1 changed files
  • Comments generated: 1
  • Review effort level: Lite

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

Comment on lines 56 to 58
- name: 'Update client service'
if: steps.changes.outputs.client == 'true'
if: steps.changes.outputs.client == 'true' && !cancelled()
uses: fjogeleit/http-request-action@v1
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants