Skip to content

Pull deps image on ARM when unchanged, add push retry#63

Merged
harshavemula-ua merged 2 commits intoCIROH-UA:mainfrom
harshavemula-ua:upgrade-arm-ci-instance
Feb 20, 2026
Merged

Pull deps image on ARM when unchanged, add push retry#63
harshavemula-ua merged 2 commits intoCIROH-UA:mainfrom
harshavemula-ua:upgrade-arm-ci-instance

Conversation

@harshavemula-ua
Copy link
Collaborator

@harshavemula-ua harshavemula-ua commented Feb 20, 2026

Summary

  • When forcingprocessor-deps hasn't changed, pull the existing image from Docker Hub instead of rebuilding from scratch on ARM
  • Adds retry logic (3 attempts) to docker_loginNpush.sh to handle transient Docker daemon failures

Context

When only forcingprocessor version changed (not deps), the workflow deleted the deps build command from the ARM execution. But since forcingprocessor uses FROM awiciroh/forcingprocessor-deps as its base, docker compose rebuilt deps from scratch anyway on the fresh EC2 instance — taking ~1 hour and exhausting resources, causing the Docker socket EOF during push (run 22201325622).

Now the workflow substitutes the deps build with docker pull awiciroh/forcingprocessor-deps:latest-arm64 so the base image is available without rebuilding.

Files changed

  • .github/workflows/build_push_fp.yaml — replace deps build delete with docker pull
  • docker/docker_loginNpush.sh — retry logic for push failures

Test plan

  • Re-run build_push_fp.yaml workflow with only fp version changed
  • Verify ARM build pulls deps image instead of rebuilding
  • Verify manifest creation step completes without errors

t4g.large was causing Docker daemon crashes due to resource exhaustion
during ARM builds, leading to failed image pushes (EOF on docker socket).
@harshavemula-ua harshavemula-ua changed the title Upgrade ARM CI instances to m8g.xlarge with push retry Pull deps image on ARM when unchanged, add push retry Feb 20, 2026
@harshavemula-ua harshavemula-ua force-pushed the upgrade-arm-ci-instance branch 3 times, most recently from 73a65d8 to 3034902 Compare February 20, 2026 16:24
@arpita0911patel arpita0911patel requested review from Copilot and removed request for arpita0911patel February 20, 2026 17:13
Copy link
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 ARM build/push execution configuration and the Docker push helper script to improve reliability of image pushes and adjust ARM builder capacity.

Changes:

  • Add retry logic (up to 3 attempts) to Docker image tag/push operations in docker_loginNpush.sh.
  • Increase ARM EC2 instance type used by ARM push/test executions from t4g.large to m8g.xlarge.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 4 comments.

File Description
docker/docker_loginNpush.sh Adds push retry loop for transient failures during tag/push operations.
.github/executions/fp_push_execution_arm.json Bumps ARM builder instance size for push execution.
.github/executions/fp_ds_test_execution_arm.json Bumps ARM builder instance size for DS test execution.

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

@harshavemula-ua harshavemula-ua merged commit fcdbb11 into CIROH-UA:main Feb 20, 2026
4 of 8 checks passed
@harshavemula-ua harshavemula-ua deleted the upgrade-arm-ci-instance branch February 20, 2026 19:27
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.

4 participants