Promote AdoptOpenJDK Version 11.0.8+10 (GA)

[gdams]

Java Version:
11.0.8+10

JVM:

• OpenJ9
• Hotspot

@AdoptOpenJDK/tsc please can somebody +1 this request?

[gdams]

TRSS Link: https://trss.adoptopenjdk.net/buildDetail?parentId=5f0ec5177eaab3443ca8322c&testSummaryResult=failed&buildNameRegex=%5ETest.*

Test Failures - Hotspot

Test_openjdk11_hs_extended.system_ppc64_aix
MathLoadTest_all_0 => long term failure #132 (comment)
LockingLoadTest_0 => long term failure #132 (comment)

Test_openjdk11_hs_extended.system_x86-64_mac
HCRLateAttachWorkload_previewEnabled_0 => long term failure adoptium/aqa-tests#1555

Test_openjdk11_hs_sanity.external_x86-64_linux_openliberty-mp-tck
openliberty_microprofile_tck_0 => long term failure #152 (comment)

[karianna]

I approve 11.0.8+10 release for Hotspot builds.

[smlambert]

As those failures have been carried over across from several releases, I would like to see them investigated/understood and resolved in time for next release.

Can you create issues for MathLoadTest_all / LockingLoadTest and put on the top priority list?

adoptium/infrastructure#1274 which is ultimately the issue for HCRLateAttachWorkload_previewEnabled failure, is already marked as top priority.

[smlambert]

OpenJ9 jdk-11.0.8+10_openj9-0.21.0

EA builds for aarch64 (pipeline 1086 and 1088):
Test_openjdk11_j9_sanity.openjdk_aarch64_linux - 1 failure, jdk_util / ReduceTest (appears 'new', crash then test times out, console output) - EA release, so non-blocking (noted here: eclipse-openj9/openj9#10168 (comment))

Test pipelines on ppc64le and ppc64le_xl failed to download jtreg (issue opened here: adoptium/aqa-tests#1894), rerun of test pipelines here:
Test_openjdk11_j9_sanity.openjdk_ppc64le_linux 293 - all pass
Test_openjdk11_j9_sanity.openjdk_ppc64le_linux_xl 192 - all pass

Test_openjdk11_j9_sanity.openjdk_ppc64_aix - jdk_util_j9_0 / Zip64SizeTest failure - most definitely a machine issue, only reproducible on 1 AIX machine: adoptium/aqa-tests#1879

Disabling build-osuosl-aix71-ppc64-1 to rerun sanity.system target in https://ci.adoptopenjdk.net/job/Test_openjdk11_j9_sanity.system_ppc64_aix/10 (Update: all pass), since all tests fail with message:

[2020-07-15T17:29:59.332Z] There must be at least 3Gb (3072Mb) free to be sure of capturing diagnostics
[2020-07-15T17:29:59.332Z] files in the event of a test failure.

Test_openjdk11_j9_sanity.system_x86-64_linux - 7 jlm targets fail (TestJlmRemoteMemoryAuth_0, TestIBMJlmRemoteMemoryAuth_0, TestJlmRemoteMemoryNoAuth_0, TestIBMJlmRemoteMemoryNoAuth_0, TestIBMJlmRemoteClassAuth_0, TestJlmRemoteClassNoAuth_0, and TestIBMJlmRemoteClassNoAuth_0) - seen before, appears to be a network issue on the machine (see eclipse-openj9/openj9#10156), will rerun here: https://ci.adoptopenjdk.net/job/Test_openjdk11_j9_sanity.system_x86-64_linux/289

Test_openjdk11_j9_sanity.openjdk_x86-64_mac_xl - jdk_util / TestCleaner - known intermittent issue considered non-blocking, eclipse-openj9/openj9#8872

[karianna]

As those failures have been carried over across from several releases, I would like to see them investigated/understood and resolved in time for next release.

Can you create issues for MathLoadTest_all / LockingLoadTest and put on the top priority list?

AdoptOpenJDK/openjdk-infrastructure#1274 which is ultimately the issue for HCRLateAttachWorkload_previewEnabled failure, is already marked as top priority.

@d3r3kk was kindly on the case for capturing the other failing tests. They're not P0/P1's but hopefully, with our extra internal infra we can now start splitting out infra vs 'real' failures on these

[lumpfish]

Test_openjdk11_j9_sanity.system_x86-64_linux rerun (https://ci.adoptopenjdk.net/job/Test_openjdk11_j9_sanity.system_x86-64_linux/289) - all tests passed.

[sxa]

Approved for OpenJ9 as per slack discussion