RHOAIENG-18459: chore(tests/containers/workbenches): make the ipv6 listening test work on macOS #868

jiridanek · 2025-01-27T10:03:30Z

follow-up for

RHOAIENG-18459: chore(tests/containers/workbenches): listen on single-stack IPv6 #866

Description

Uses the pydantic library to load output from podman machine info and sets-up a ssh tunnel from the podman machine vm to the machine running the test.

This is a workaround for macOS/rootful issues in Podman

How Has This Been Tested?

works on my macOS
https://github.com/jiridanek/notebooks/actions/runs/12986604939

poetry run pytest tests/containers \
    -k 'TestWorkbenchImage and test_ipv6_only' \
    --image=quay.io/opendatahub/workbench-images:codeserver-ubi9-python-3.11-20250124-6a93e34

Merge criteria:

The commits are squashed in a cohesive manner and have meaningful messages.
Testing instructions have been added in the PR body (for PRs involving changes that are not immediately obvious).
The developer has manually tested the changes and verified that the changes work

…stening test work on macOS

jiridanek · 2025-01-27T10:42:24Z

tests/containers/workbenches/workbench_image_test.py

/override ci/prow/images
n/a

/override ci/prow/images
this is my override thread

openshift-ci · 2025-01-27T10:42:41Z

@jiridanek: Overrode contexts on behalf of jiridanek: ci/prow/images

In response to this:

/override ci/prow/images
n/a

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

jstourac · 2025-01-27T11:30:29Z

/lgtm

andyatmiami · 2025-01-27T13:47:32Z

tests/containers/podman_machine_utils.py

+# Usage example
+if __name__ == "__main__":
+    tunnel_process = open_ssh_tunnel(
+        "podman-machine-default",


not sure of a good/right place to document this.. but we should ideally call out (for purposes of running locally) - we require a machine named podman-machine-default to exist. I realize it errors after the fact appropriately in this script... but this is codifying a requirement worth calling out in a readme or something maybe?

(as a case in point - I don't (as of now - but will soon!) have a podman-machine-default machine - as its too underpowered to run many notebooks)

so, you have a differently named podman machine? in that case this should probably detect whatever machine you're using and use that, wdyt?

yes, perhaps if we just had the system choose the currently active connection - that makes it "one less stumbling block"

➜ ~/ $ podman system connection ls Name URI Identity Default ReadWrite rhoai-default ssh://core@127.0.0.1:50943/run/user/501/podman/podman.sock /Users/astonebe/.local/share/containers/podman/machine/machine false true rhoai-default-root ssh://root@127.0.0.1:50943/run/podman/podman.sock /Users/astonebe/.local/share/containers/podman/machine/machine true true

seems podman system connection ls can give you a data structure that lists which machine is currently default

very good point, I added a commit here that addresses that; I look what the realpath docker_socket is, and compare that with the sockets given in podman machine info. I did not need podman system connections ls after all

And, turned out this does not really fix anything with 1+ podman machines present. That's because podman machine inspect without machine name inspects the default machine only. Followup pr needed.

andyatmiami · 2025-01-27T13:49:30Z

tests/containers/schemas.py

@@ -0,0 +1,41 @@
+from pydantic import BaseModel


Did you look into https://podman-py.readthedocs.io/en/stable/index.html ? Can we get some of this boilerplate "for free" if we use the Podman Python SDK? (or, conversely, is the Podman Python SDK "more hassle than its worth" ?)

there's quite a lot of various docker/podman libraries like this

since I went with testcontainers (which uses docker-py) originally, I intended to keep to docker-compatible apis; but, turns out, more is oftentimes needed

besides testcontainers, there's all that Adriel mentioned at one of the qe meetings

https://terratest.gruntwork.io/examples/

https://github.com/pytest-dev/pytest-testinfra

https://dagger.io/

there are other stuff, for example something from red hat people that runs podman commands as a subprocess; so whatever you script in python, you have an obvious way to reproduce on commandline

I don't think that podman machine api is exposed there... but I'd have to check more thoroughly to find out.

I spent few minutes and no joy so far. I really think podman-machine not part of the libpod api.

totally fine - not worth spending much time on it - appreciate you sharing thoughts/motivation behind the question... you can consider this conversation resolved.

it's a good thought, cc @jstourac maybe we'd just stop pretending we use docker and will fully embrace podman

there's also https://issues.redhat.com/browse/RHOAIENG-18060 that asks to write these test so that they can be run on docker/podman and also openshift; that's gonna be another challenge, but maybe a lesser one than all this macOS stuff

…dling: detect the right one to use if user has multiple machines around

openshift-ci · 2025-01-27T15:52:21Z

@jiridanek: Overrode contexts on behalf of jiridanek: ci/prow/images

In response to this:

/override ci/prow/images
this is my override thread

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

daniellutz · 2025-01-27T16:08:32Z

I tested here and looks good!

/lgtm

andyatmiami · 2025-01-27T16:19:57Z

/lgtm

I have confirmed if I have a machine named podman-machine-default - the given test command succeeds:

poetry run pytest tests/containers \
    -k 'TestWorkbenchImage and test_ipv6_only' \
    --image=quay.io/opendatahub/workbench-images:codeserver-ubi9-python-3.11-20250124-6a93e34

...
PASSEDDEBUG:urllib3.connectionpool:http://localhost:None "DELETE /v1.41/networks/e252ab284eaa64991cb4b40e68e99cd337f9723544d308caadd7de65b0a6c215 HTTP/1.1" 204 0


=================================================================================================== 1 passed, 4 deselected in 199.02s (0:03:19) ===================================================================================================

ℹ️ I am on MacOS

jiridanek · 2025-01-27T16:30:25Z

/approve

openshift-ci · 2025-01-27T16:30:32Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: jiridanek

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

~~OWNERS~~ [jiridanek]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

openshift-ci bot requested review from caponetto and harshad16 January 27, 2025 10:03

openshift-ci bot added the size/l label Jan 27, 2025

jiridanek force-pushed the jd_initial_ipv6_w_macos branch from c3fb7f2 to 36045ab Compare January 27, 2025 10:04

openshift-ci bot added size/l and removed size/l labels Jan 27, 2025

RHOAIENG-18459: chore(tests/containers/workbenches): make the ipv6 li…

d77fbda

…stening test work on macOS

jiridanek force-pushed the jd_initial_ipv6_w_macos branch from 36045ab to d77fbda Compare January 27, 2025 10:10

openshift-ci bot added size/l and removed size/l labels Jan 27, 2025

jiridanek added the tide/merge-method-squash Denotes a PR that should be squashed by tide when it merges. label Jan 27, 2025

jiridanek commented Jan 27, 2025

View reviewed changes

openshift-ci bot assigned jstourac Jan 27, 2025

openshift-ci bot added lgtm size/l and removed size/l labels Jan 27, 2025

andyatmiami reviewed Jan 27, 2025

View reviewed changes

RHOAIENG-18459: chore(tests/containers): improve podman machine han…

8f66506

…dling: detect the right one to use if user has multiple machines around

openshift-ci bot added size/l and removed lgtm size/l labels Jan 27, 2025

jiridanek added tide/merge-method-rebase Denotes a PR that should be rebased by tide when it merges. and removed tide/merge-method-squash Denotes a PR that should be squashed by tide when it merges. labels Jan 27, 2025

openshift-ci bot assigned daniellutz Jan 27, 2025

openshift-ci bot added the lgtm label Jan 27, 2025

openshift-ci bot assigned andyatmiami Jan 27, 2025

openshift-ci bot added the approved label Jan 27, 2025

openshift-merge-bot bot merged commit 9e76f6f into opendatahub-io:main Jan 27, 2025
7 checks passed

jiridanek deleted the jd_initial_ipv6_w_macos branch January 27, 2025 16:54

jiridanek mentioned this pull request Jan 27, 2025

RHOAIENG-18459: chore(tests/containers): handle multiple podman machines and improve debugging output when podman machine is not found #869

Merged

3 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

RHOAIENG-18459: chore(tests/containers/workbenches): make the ipv6 listening test work on macOS #868

RHOAIENG-18459: chore(tests/containers/workbenches): make the ipv6 listening test work on macOS #868

jiridanek commented Jan 27, 2025 •

edited

Loading

jiridanek Jan 27, 2025

jiridanek Jan 27, 2025

openshift-ci bot commented Jan 27, 2025

jstourac commented Jan 27, 2025

andyatmiami Jan 27, 2025

jiridanek Jan 27, 2025

andyatmiami Jan 27, 2025

jiridanek Jan 27, 2025

jiridanek Jan 28, 2025

andyatmiami Jan 27, 2025

jiridanek Jan 27, 2025

jiridanek Jan 27, 2025

andyatmiami Jan 27, 2025

jiridanek Jan 27, 2025

openshift-ci bot commented Jan 27, 2025

daniellutz commented Jan 27, 2025

andyatmiami commented Jan 27, 2025 •

edited

Loading

jiridanek commented Jan 27, 2025

openshift-ci bot commented Jan 27, 2025

RHOAIENG-18459: chore(tests/containers/workbenches): make the ipv6 listening test work on macOS #868

RHOAIENG-18459: chore(tests/containers/workbenches): make the ipv6 listening test work on macOS #868

Conversation

jiridanek commented Jan 27, 2025 • edited Loading

Description

How Has This Been Tested?

Merge criteria:

Choose a reason for hiding this comment

Choose a reason for hiding this comment

openshift-ci bot commented Jan 27, 2025

jstourac commented Jan 27, 2025

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

openshift-ci bot commented Jan 27, 2025

daniellutz commented Jan 27, 2025

andyatmiami commented Jan 27, 2025 • edited Loading

jiridanek commented Jan 27, 2025

openshift-ci bot commented Jan 27, 2025

jiridanek commented Jan 27, 2025 •

edited

Loading

andyatmiami commented Jan 27, 2025 •

edited

Loading