chore: Stabilize Valkey and Mongodb tests - #334
Merged
Merged
Conversation
Gu1nness
approved these changes
Sep 25, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Valkey
The test deployed
valkeyfrom9/edgewithout a base. The channel holds one revision per base: rev 88 for ubuntu@24.04 and rev 99 for ubuntu@26.04. Juju 3.6 uses its default base when none is given, so the test got rev 88, an old 24.04 build. The Valkey charm has moved to ubuntu@26.04.The test now passes
--base=ubuntu@26.04and gets rev 99. It deploys Valkey withops_test.juju("deploy", ...)because python-libjuju cannot do this:model.deploy(..., base="ubuntu@26.04")resolves the revision before it applies the base. It picks rev 88 and fails withbase "ubuntu@26.04/stable" is not supported, supported bases are: ubuntu@24.04.revision=99fails withUnknown series: resolute, because libjuju's series table stops at 24.04.I checked each case against a local Juju 3.6.28 controller, the same version CI uses.
A valkey bug is being fixed where
storage-attachedcrashes when the workload container's Pebble socket does not exist yet. The fix is in canonical/valkey-operator.MongoDB
mongodb-k8sreportsactivebefore itsmongodcontainer is up. The test waited for that status, related data-integrator, and assertedactiveafter a 15s idle window. MongoDB skips client requests until it initialises the database, so data-integrator stayed blocked on "Please relate the data-integrator with the desired product".This test used to pass because workload containers started faster in CI. In the last green run,
pebble-readyfired about 20s after the unit started. In the failing runs it took about 60s, and in the MongoDB job it never fired before the assert.MongoDB creates users for existing relations once it initialises the database. The test now waits up to 1000s for both apps to reach
active, the same waytest_valkey.pywaits.Changes
tests/integration/constants.py: addVALKEY_BASE = "ubuntu@26.04".tests/integration/test_valkey.py: deploy Valkey through the Juju CLI with--base.tests/integration/test_mongodb.py: after relating, wait forstatus="active"in place of the immediate assert.