Skip to content

WIP: e2e-tests: share common first steps via VM checkpoints - #1714

Draft
adombeck wants to merge 4 commits into
mainfrom
UDENG-10153-e2e-test-snapshots
Draft

WIP: e2e-tests: share common first steps via VM checkpoints#1714
adombeck wants to merge 4 commits into
mainfrom
UDENG-10153-e2e-test-snapshots

Conversation

@adombeck

@adombeck adombeck commented Jul 9, 2026

Copy link
Copy Markdown
Contributor

No description provided.

@adombeck adombeck added the e2e-tests This issue is related to end-to-end tests / Run end-to-end tests on this pull request label Jul 9, 2026
@adombeck
adombeck force-pushed the UDENG-10153-e2e-test-snapshots branch from df05c8b to 3cdfa4b Compare July 9, 2026 15:46
adombeck and others added 4 commits July 18, 2026 16:09
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
When a test's setup finds that the shared checkpoint it depends on
previously failed (or failed concurrently), the test itself cannot
meaningfully run.  Reporting it as FAIL conflates a genuine test
failure with a dependency problem.

Tag the test with `checkpoint-failed` before raising in the two
Checkpoint.setup_checkpoint paths that surface a prior failure (the
"previously failed" and "concurrent failure" cases).  The listener
then rewrites any FAIL result that carries this tag to NOT RUN in
end_test, so those tests are visually and statistically distinct from
both genuine failures and broker-not-applicable skips.

The NOT RUN case is also added to the listener's status-colour table,
and the fixed-width :4 format spec on the status string is dropped so
that "NOT RUN" aligns correctly alongside the shorter status labels.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
When Journal.Start Receiving Journal connects to the VM via TCP (port
55000), a virsh memory snapshot taken while that connection is alive
captures the TCP state as ESTABLISHED.  The first test that restores
the checkpoint gets a VM whose TCP stack still considers the old socat
connection open; new socat connections are refused until the VM's TCP
stack times the stale entry out (~60 s).

Fix this by stopping the journal and VNC recording before calling
Snapshot.Create, so the checkpoint snapshot is taken with no active
connections on port 55000.  Journal and recording are restarted
immediately after the snapshot so the creator test's body still has
them available.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
@adombeck
adombeck force-pushed the UDENG-10153-e2e-test-snapshots branch from 3cdfa4b to fe2970e Compare July 18, 2026 14:09
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

e2e-tests This issue is related to end-to-end tests / Run end-to-end tests on this pull request

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant