Skip to content

improve CI build times (WIP)#2322

Open
peterj wants to merge 1 commit into
mainfrom
peterj/cimprove
Open

improve CI build times (WIP)#2322
peterj wants to merge 1 commit into
mainfrom
peterj/cimprove

Conversation

@peterj

@peterj peterj commented Jul 23, 2026

Copy link
Copy Markdown
Collaborator

wip

Copilot AI review requested due to automatic review settings July 23, 2026 14:21
@peterj
peterj requested a review from Charlesthebird as a code owner July 23, 2026 14:21

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Pull request overview

This PR aims to reduce CI build time by improving Docker BuildKit layer cache reuse/avoidance across jobs and by narrowing multi-arch validation on pull requests.

Changes:

  • Added CI-controlled per-image BuildKit type=gha cache arguments in the root Makefile.
  • Updated CI workflow to use the new cache variables and to build only linux/amd64 on PRs (both amd64/arm64 on main/release pushes).
  • Removed the “cancel previous actions” step from the shared initialize-environment composite action.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File Description
Makefile Introduces docker-cache-args and wires it into image build targets to enable per-image GHA cache scopes.
.github/workflows/ci.yaml Switches cache configuration to CI_CACHE_* env vars, adjusts platforms per event type, and trims the build matrix.
.github/actions/initialize-environment/action.yaml Removes a workflow-cancel step from the shared init action.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment thread Makefile
Comment thread .github/workflows/ci.yaml
@peterj
peterj force-pushed the peterj/cimprove branch 3 times, most recently from b0978a2 to 8eac06d Compare July 24, 2026 12:13
Signed-off-by: Peter Jausovec <peter.jausovec@solo.io>
@peterj
peterj force-pushed the peterj/cimprove branch from 8eac06d to 92af71b Compare July 24, 2026 14:51
Comment thread Makefile
ACP_SANDBOX_HERMES_IMG ?= $(DOCKER_REGISTRY)/$(DOCKER_REPO)/$(ACP_SANDBOX_HERMES_IMAGE_NAME):$(ACP_SANDBOX_IMAGE_TAG)
ACP_SANDBOX_OPENCLAW_IMG ?= $(DOCKER_REGISTRY)/$(DOCKER_REPO)/$(ACP_SANDBOX_OPENCLAW_IMAGE_NAME):$(ACP_SANDBOX_IMAGE_TAG)
ACP_SANDBOX_CLAUDE_IMG ?= $(DOCKER_REGISTRY)/$(DOCKER_REPO)/$(ACP_SANDBOX_CLAUDE_IMAGE_NAME):$(ACP_SANDBOX_IMAGE_TAG)
E2E_ACP_SANDBOX_OPENCLAW_IMG ?= $(ACP_SANDBOX_OPENCLAW_IMG)@sha256:eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

hm, these fake sha256:eeee digests get baked straight into AcpSandboxOpenClawImageDigest/HermesImageDigest via the ldflags script (it shortcuts on the @sha256: form and skips resolving a real one), and constants.go uses those to build the actual image ref, so the day sth adds an e2e test w openclaw or hermes runtime it'll just imagepullbackoff w no clue why

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