Weekly build triage for the week starting Saturday 2023/06/24

[adamfarley]

This week's build triage summary:

Date	Day	JDK Version	Pipeline	Pass/Build Fail/Test Fail	Pass/Preventable/Unpreventable
2023/06/24	Saturday	JDK8	Link	9/1/1	9/1/1
2023/06/24	Saturday	JDK11	Link	9/2/0	9/2/0
2023/06/25	Sunday	JDK17	Link	10/0/1	10/1/0
2023/06/25	Sunday	JDK20	Link	9/0/1	9/0/1
2023/06/25	Sunday	JDK21	Link	9/0/1	9/0/1

Note: "Test Fail" is for when all the "build" jobs passed (build, sign, installer, etc) but one of the test jobs failed and the
status propagated upstream to the build job. Note that "unstable" test job status can be considered a pass, as these are
triaged in more detail here.

Comment template:

Triage breakdown for <Weekday>

List of failures:

<link to failing job>
<Problem summary>
New/Existing Issue: <Issue link>
Preventable: Yes/No

Repeat as needed.

[adamfarley]

Triage breakdown for Weekend

List of failures:

https://ci.adoptium.net/job/build-scripts/job/jobs/job/jdk8u/job/jdk8u-aix-ppc64-temurin/312/console
JDK7 boot directory environment variable on jenkins is set to the location of an IBM jdk that can't build Temurin.
Also, there's nothing at that location, so it can't find the JDK that would have failed to build it.
Double failure?
New problem, Existing Issue: #3261 (comment)
Preventable: Yes

https://ci.adoptium.net/job/Test_openjdk8_hs_extended.openjdk_sparcv9_solaris/94/
Timeout. I thought this was solved. Apparently not. Also, some/all of the excluded tests are still being run. Hrmmm.
Existing Issue: adoptium/infrastructure#2893 (comment)
Preventable: No

https://ci.adoptium.net/job/build-scripts/job/jobs/job/jdk11u/job/jdk11u-aix-ppc64-temurin/255/
TOC overflow, followed by a series of "*.o is not valid with the current object file mode." errors.
It's possible this issue was caused by an environment variable not being populated.
The infrastructure team has been notified.
Existing Issue: #2734 (comment)
Preventable: Yes

https://ci.adoptium.net/job/Test_openjdk11_hs_sanity.external_x86-64_linux/590/
No space left on device.
New Issue: adoptium/infrastructure#3117
Preventable: Yes

https://ci.adoptium.net/job/Test_openjdk17_hs_extended.openjdk_ppc64_aix/99/
java.net.UnknownHostException: adopt02: adopt02: Hostname and service name not provided or found
Looks like an incomplete reprovisioning. The /etc/hosts file thinks the machine is "adopt10", while the uname output thinks the machine is "adopt02".
New problem, Existing Issue: adoptium/infrastructure#3030 (comment)
Preventable: Yes

https://ci.adoptium.net/job/Test_openjdk20_hs_extended.openjdk_aarch64_mac/41/
test-macstadium-macos11-arm64-2 was marked offline
Looks like a networking issue. Ignoring until this becomes a pattern.
Preventable: No

https://ci.adoptium.net/job/Test_openjdk21_hs_extended.openjdk_x86-64_windows/14/
test-azure-win2019-x64-1 appears to be running very slowly. Tests that pass appear to be 5x slower than tests elsewhere (alibaba example) on the same jdk version and platform.
Update: Seems only a hair slower than other azure machines though (example). Seems the azure machines have been running slowly for a while now. They've been completing this test set within minutes of the 10hr timeout. Is something choking the azure machines' resources, or could they simply be slow. Not sure. Raising an infrastructure issue to find out.
New/Existing Issue: tbd
Preventable: No