Both attached storage and networking in the EC2 environment utilized by CI are slower than might be experienced on a local machine. Increasing allocated EC2 bandwidth to either comes at a hefty cost. But more importantly, one cannot predict the capabilities of a local machine (i.e. network hiccups, attached slow USB storage, etc.). So testing w/ a reasonable lowest-common-denominator is an overall net-benefit, if the tests can be made to pass.
IMPORTANT
When this issue is resolved, please revert the Do not aggregate failing mac test status commit added in PR #20691
Issue Description
Both attached storage and networking in the EC2 environment utilized by CI are slower than might be experienced on a local machine. Increasing allocated EC2 bandwidth to either comes at a hefty cost. But more importantly, one cannot predict the capabilities of a local machine (i.e. network hiccups, attached slow USB storage, etc.). So testing w/ a reasonable lowest-common-denominator is an overall net-benefit, if the tests can be made to pass.
IMPORTANT
When this issue is resolved, please revert the
Do not aggregate failing mac test status
commit added in PR #20691Steps to reproduce the issue
Steps to reproduce the issue
make podman-remote
make localmachine
Describe the results you received
The basic tests all pass (in ~15 minutes). The full suite mostly fails
Describe the results you expected
Both basic and full suite CI tests should all pass.
podman info output
Podman in a container
No
Privileged Or Rootless
Rootless
Upstream Latest Release
Yes
Additional environment details
Tests running as a regular non-admin user. Using:
mac2.metal
EC2 instancegp3
volume w/ 300 MB/s allocated IOPS14.1-20231110-071234-arm64
Additional information
No response