Cloning from GitHub randomly timed out resulting in failed tests

[jackorp]

Container platform

No response

Version

No response

OS version of the container image

No response

Bugzilla, Jira

No response

Description

In recent nightly tests, Ruby containers failed on RHEL 8 and RHEL 9.

The problem is git clone timing out. This seems like random/intermittent issue.

It happened with 2 different Ruby versions, in different tests cloning different repos, so it shouldn't be an issue with the container itself.

1 failure was with Ruby 2.5 'db' run_s2i_build' https://github.com/openshift/ruby-hello-world.git/ , and second was with Ruby 3.1 'from_dockerfile' test_from_dockerfile_s2i pulling rails-ex https://github.com/sclorg/rails-ex.git/ .

Rest of the Ruby container test suite for other version and where other cloning happens finished OK.

Trying to clone in a loop to should prevent an intermittent timeout failing the tests.

Relevant excerpts from logs:
RHEL9 Ruby 'from_dockerfile' test_from_dockerfile_s2i:

fatal: unable to access 'https://github.com/sclorg/rails-ex.git/': Failed to connect to [github.com](http://github.com/) port 443: Connection timed out
ERROR: Git repository https://github.com/sclorg/rails-ex.git@master cannot be cloned into app-src.
Terminating the Dockerfile build.
Test for image 'ubi9/ruby-31:1' FAILED (exit code: 1)
Test for image 'ubi9/ruby-31:1' FAILED (exit code: 1)

RHEL 8 Ruby db' run_s2i_build:

fatal: unable to access 'https://github.com/openshift/ruby-hello-world.git/': Failed to connect to [github.com](http://github.com/) port 443: Connection timed out
/var/tmp/daily_scl_tests/rhel8-test/plans/nightly-container-rhel8/data/s2i-ruby-container/2.5
cp: cannot stat '/var/tmp/daily_scl_tests/rhel8-test/plans/nightly-container-rhel8/data/s2i-ruby-container/2.5/test/db-test-app/.': No such file or directory

Reproducer

No response