Skip to content

Add integration test for upgrading from edge#4557

Merged
olix0r merged 12 commits intomasterfrom
kleimkuhler/add-upgrade-edge
Jun 16, 2020
Merged

Add integration test for upgrading from edge#4557
olix0r merged 12 commits intomasterfrom
kleimkuhler/add-upgrade-edge

Conversation

@kleimkuhler
Copy link
Contributor

@kleimkuhler kleimkuhler commented Jun 4, 2020

Problem

#4557 changed the name of the function that helm_upgrade_integration_tests
uses.

install_stable() was renamed to latest_release_channel() and now takes an
argument for specifying either edge or stable.

run_helm_upgrade_test is a function used by the helm upgrade integration test
and was not properly updated to use latest_release_channel().

This silently passed integration tests because run_helm_upgrade_test started
passing an empty string for the version to upgrade from, which results in the
default behavior of install_test.go--and therefore still passes.

Solution

run_helm_upgrade_test now uses latest_release_channel() and passes the
proper argument.

Additionally, it checks that the version returned from
latest_release_channel() is not empty. If it is empty, it exits the test. This
ensures something like this does happen in the future.

Signed-off-by: Kevin Leimkuhler [email protected]

@kleimkuhler kleimkuhler requested a review from alpeb as a code owner June 4, 2020 22:25
Signed-off-by: Kevin Leimkuhler <[email protected]>
Signed-off-by: Kevin Leimkuhler <[email protected]>
Copy link
Member

@alpeb alpeb left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks @kleimkuhler. It seems this requires some sort of Github admin setting to be tweaked in order to run?
Note that this also fixes #4472

Signed-off-by: Kevin Leimkuhler <[email protected]>
Signed-off-by: Kevin Leimkuhler <[email protected]>
Signed-off-by: Kevin Leimkuhler <[email protected]>
Signed-off-by: Kevin Leimkuhler <[email protected]>
@alpeb
Copy link
Member

alpeb commented Jun 8, 2020

This might depend on #4571 in order to pass, so I'm merging that one now.

Signed-off-by: Kevin Leimkuhler <[email protected]>
Copy link
Member

@alpeb alpeb left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM 👍

Signed-off-by: Kevin Leimkuhler <[email protected]>
Copy link
Contributor

@Pothulapati Pothulapati left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM!

@grampelberg
Copy link
Contributor

@kleimkuhler what do we need to do to get this merged?

@kleimkuhler
Copy link
Contributor Author

@grampelberg It needs to be admin merged and then the expected checks need to be updated in GitHub settings.

This PR changes a check name from Integration tests (upgrade) to Integration test (upgrade_stable). All checks have passed, it just expects a test that is changed by this PR.

@grampelberg
Copy link
Contributor

@admc can you help out?

@olix0r olix0r merged commit b0765c4 into master Jun 16, 2020
@olix0r olix0r deleted the kleimkuhler/add-upgrade-edge branch June 16, 2020 16:18
kleimkuhler added a commit that referenced this pull request Jun 18, 2020
## Problem

#4557 changed the name of the function that `helm_upgrade_integration_tests`
uses.

`install_stable()` was renamed to `latest_release_channel()` and now takes an
argument for specifying either `edge` or `stable`.

`run_helm_upgrade_test` is a function used by the helm upgrade integration test
and was not properly updated to use `latest_release_channel()`.

This silently passed integration tests because `run_helm_upgrade_test` started
passing an empty string for the version to upgrade from, which results in the
default behavior of `install_test.go`--and therefore still passes.

## Solution

`run_helm_upgrade_test` now uses `latest_release_channel()` and passes the
proper argument.

Additionally, it checks that the version returned from
`latest_release_channel()` is not empty. If it is empty, it exits the test. This
ensures something like this does happen in the future.

Signed-off-by: Kevin Leimkuhler <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants