upgrade core LTS to 2.346.1 and associated bom, fix test due to this upgrade

[olamy]

Signed-off-by: Olivier Lamy olamy@apache.org

[olamy]

I have created the label it-works-on-my-machine (^&%&%$(&**&(%&* :))

[olamy]

interesting CI fail because the record logs are different in Java 8 and Java 11.

Java 8
[
Generating App Installation Token for app ID 54321, 
Generating App Installation Token for app ID 54321, 
Failed to generate new GitHub App Installation Token for app ID 54321: cached token is stale but has not expired, 
Generating App Installation Token for app ID 54321, 
Failed to generate new GitHub App Installation Token for app ID 54321: cached token is stale but has not expired, 
Generating App Installation Token for app ID 54321, 
Generating App Installation Token for app ID 54321]


Java 11
[Generating App Installation Token for app ID 54321, 
Generating App Installation Token for app ID 54321, 
Failed to generate new GitHub App Installation Token for app ID 54321: cached token is stale but has not expired, 
Generating App Installation Token for app ID 54321, 
Failed to generate new GitHub App Installation Token for app ID 54321: cached token is stale but has not expired, 
Generating App Installation Token for app ID 54321]

Java 8 has one more line.

[olamy]

I'm tempted to ignore Java 8 as anyway Jenkins has already moved to Java 11.

[olamy]

now doesn't work on windows with Java 8....

[jtnord]

so is it time to declare the test flaky yet or does it reliably fail somewhere?

[olamy]

so is it time to declare the test flaky yet or does it reliably fail somewhere?

with the upgrade it's stable with linux/java11 and windows/java8.
but really flaky with linux/java8...

As we will probably move soon to java11 this will not be a problem.
maybe we can simply skip this test for linux/java8 until we upgrade?

[jtnord]

Ahh..

@olamy this is incorrect - the test was failing due to JENKINS-69488 which is why my PR is breaking this test :)

[olamy]

good catch