Closed sicoyle closed 1 year ago
Hey @mukundansundar! Do you have any pointers on e2e tests for v1.11.0 features? I added the external service invocation functionality with the new HTTPEndpoint Dapr resource, and am having a hard time with configuring the e2e tests for it.
Does the runtime v1.11.0 release have to be cut before those would pass? I feel like this is almost a chicken before the egg scenario here... I saw the CRDs work off the raw data for the specific runtime version yaml files so I tried a work around pointing to master for the HTTPEndpoint yaml file, but then it tried to look for a "master" helm-chart release which does not exist. I also see TODO comments in common.go
stating TODO: Skip the dashboard version check for now until the helm chart is updated
, so that had me thinking that I also need the updated helm chart.
This will depend on latest runtime changes and will need runtime RC at-least.
This will depend on latest runtime changes and will need runtime RC at-least.
Thank you @pravinpushkar!! I will circle back here then when the release is ready!
@sicoyle dapr RC is out. Probably this PR can make use of that.
@sicoyle dapr RC is out. Probably this PR can make use of that.
I think 1284 needs to be merged before that.
Merging #1280 (d44d942) into master (36a13bc) will not change coverage. The diff coverage is
n/a
.:exclamation: Current head d44d942 differs from pull request most recent head e5e5522. Consider uploading reports for the commit e5e5522 to get more accurate results
@@ Coverage Diff @@
## master #1280 +/- ##
=======================================
Coverage 27.03% 27.03%
=======================================
Files 39 39
Lines 3873 3873
=======================================
Hits 1047 1047
Misses 2752 2752
Partials 74 74
E2E self hosted failure seems unrelated and due to timing out on installing podman:
Waiting for VM ...
Error: The action has timed out.
also upload test results failure shows:
Run actions/upload-artifact@master
Error: Input required and not supplied: path
I think there may be a github issue or workflow connectivity issue going on for that last build because looking through the E2E upgrade path tests I see it's failing because of timing issues again:
For example, common.go:792
is from a 2 minute timeout that expired waiting for pods to delete.
goroutine 786 [chan send, 18 minutes]:
github.com/dapr/cli/tests/e2e/common.waitPodDeletion(0xc0004ccfd0?, 0xf51f2a?, 0x0?)
/home/runner/work/cli/cli/src/github.com/dapr/cli/tests/e2e/common/common.go:1031 +0x1f1
created by github.com/dapr/cli/tests/e2e/common.uninstallTest.func1
/home/runner/work/cli/cli/src/github.com/dapr/cli/tests/e2e/common/common.go:792 +0x198
documenting again bc I've seen this issue inconsistently as well in my many e2e test failures 😅 The inconsistency part is definitely a timing related issue we should capture and address in the future; however, for the time being, this should have a conditional since the dapr dashboard is removed from the helm chart in v1.11 RC here. I am opening a PR to address separately.
=== RUN TestUpgradePathNonHAModeMTLSDisabled/vv1.11.0-rc.4_to_v1.10.5/clusterroles_exist_v1.11.0-rc.4
common.go:454:
Error Trace: /home/runner/work/cli/cli/src/github.com/dapr/cli/tests/e2e/common/common.go:454
Error: Not equal:
expected: true
actual : false
Test: TestUpgradePathNonHAModeMTLSDisabled/vv1.11.0-rc.4_to_v1.10.5/clusterroles_exist_v1.11.0-rc.4
Messages: cluster role dapr-dashboard, found = false, wanted = true
Description
Add checks for new Dapr resource httpendpoints for external service invocation as per comment here.
Issue reference
https://github.com/dapr/dapr/pull/6227
Please reference the issue this PR will close: #[4549]
Checklist
Please make sure you've completed the relevant tasks for this PR, out of the following list: