Please put the name of the software product (and affected platforms if relevant) in the title of this issue

Delete as appropriate from this list:

Bug in ansible playbook

Details:

Current playbooks do not execute out of the box on CentOS Stream 10 and will therefore likely fail on RHEL10 and other derivative distributions. It may be due to issues caused with a newer version of ansible as CS9 does not have the same problem:

0.930 + ansible-playbook -i hosts ansible/playbooks/AdoptOpenJDK_Unix_Playbook/main.yml -e git_sha=123456 --skip-tags=debug,hosts_file,hostname,adoptopenjdk,jenkins,nagios,superuser,docker,swap_file,crontab,nvidia_cuda_toolkit                                      
1.462 ERROR! couldn't resolve module/action 'homebrew'. This often indicates a misspelling, missing collection, or incorrect module path.
1.462 
1.462 The error appears to be in '/ansible/ansible/playbooks/AdoptOpenJDK_Unix_Playbook/roles/curl/tasks/main.yml': line 121, column 3, but may
1.462 be elsewhere in the file depending on the exact syntax problem.
1.462 
1.462 The offending line appears to be:
1.462 
1.462 
1.462 - name: Install latest version of curl for MacOS
1.462   ^ here
------

Feb 05 '25 15:02 sxa

Seems to be even more than this if you skip that particular part. Seen similar on RHEL9 (real installation) and there were at least three similarly problematic roles there out of the box.

Mar 18 '25 15:03 sxa

Noting that this also requires the changes described in https://github.com/adoptium/infrastructure/issues/3851 At present, running test jobs requires you to specify &&sw.os.rhel.10 onto any hostname you specify in a test job's LABEL otherwise this code from this PR will try to start Xvfb.

May 12 '25 12:05 sxa

Most of the required modules sych as homebrew, authorized_key zypper mount sysctl and alternatives seem to be provided by adding the ansible-collection-community-general package

May 12 '25 12:05 sxa

CentOS Stream 10 test results

I'm going to summarise CS10 here, but I'll also use the opportunity to summarise the CS9 runs in a follow-on comment. AQA_Test_Pipeline: https://ci.adoptium.net/job/AQA_Test_Pipeline/419/ Re-grind links are on alternate machines for comparison.

Unstable jobs:

https://ci.adoptium.net/job/Test_openjdk24_hs_sanity.openjdk_aarch64_linux/50/testReport (8 failures - some TZ related - test material mismatch likely: run-time=2025a, tests=2025b) Re-grind
https://ci.adoptium.net/job/Test_openjdk24_hs_extended.openjdk_aarch64_linux/50/testReport (202 failures) Most, but not all, display related. Weston issue Re-grind
https://ci.adoptium.net/job/Test_openjdk24_hs_dev.functional_aarch64_linux/32/testReport (6 failures) Re-grind passed SSL test failures
https://ci.adoptium.net/job/Test_openjdk24_hs_special.openjdk_aarch64_linux/32/testReport (Unclear what failed on first look) Re-grind Same issue

May 15 '25 13:05 sxa

CentOS Stream 9 test results:

Unstable jobs (Note did not run the full set from CS10 above):

https://ci.adoptium.net/job/Test_openjdk24_hs_extended.openjdk_aarch64_linux/47/testReport (4 failures) Re-grind in U18.04 4 failures
https://ci.adoptium.net/job/Test_openjdk24_hs_sanity.openjdk_aarch64_linux/48/testReport (7 failures - some TZ related as with CS10) Re-grind on another machine TIMED OUT Try again

May 15 '25 13:05 sxa

centosstream10 X11 errors are due to:

00:46:43  [19:46:42.898] fatal: environment variable XDG_RUNTIME_DIR is not set.

I have created a /home/jenkins/xdg, mode 0700 (required) and then set XDG_RUNTIME_DIR on the jenkins agent definition to that directory, and am trying again: https://ci.adoptium.net/job/Grinder/13088/console weston segfaulted - I also needed /tmp/.X11-unix to be created (as the jenkins user is ok) otherwise weston segfaults.

Running at https://ci.adoptium.net/job/Grinder/13089/

EDIT: The above test had 7 failures, so the weston/XDG fixes were successful.

May 15 '25 14:05 sxa

New runs using nightly test material for JDK21:

CentOS9

Passed with Green status 😃

CentOS10

Failed in dev.functional - six SSL-related tests. Most of those should disappear with the addition of openssl to the AQA DockerStatic image. This is pulled in as a dependency automatically for CS9 but seemingly not for CS10. New run after adding it manually to the current agent is at https://ci.adoptium.net/job/Test_openjdk21_hs_dev.functional_aarch64_linux/57/ although we are expecting one failure to remain which will be a legitimate difference between EL9 and 10.

May 19 '25 15:05 sxa

Noting that for RHEL9 you can install EPEL to get a "full" version of ansible which will allow the homebrew tasks with:

rpm --import https://dl.fedoraproject.org/pub/epel/RPM-GPG-KEY-EPEL-9
dnf install https://dl.fedoraproject.org/pub/epel/epel-release-latest-9.noarch.rpm
dnf install ansible

(Ref the RH docs, but that uses a different method for getting the EPEL key. Noting that for RHEL9 we are doing this with the rpm_key ansible task)

May 22 '25 11:05 sxa

JDK24 nightly build

CentOS9

Failures in:

sanity.openjdk (1 failure - runtime/NMT/CheckForProperDetailStackTrace.java)
special.openjdk (2 failures - maybe not important)
extended.openjdk (1 failure - same as sanity)

CentOS10

dev.functional - single failure in ssl-tests-openssl-client.sh (Likely policy failure which is EL10 specific)

May 22 '25 15:05 sxa

To install Ansible on RHEL10 to run the playbooks:

Add EPEL Repo:

dnf install https://dl.fedoraproject.org/pub/epel/epel-release-latest-10.noarch.rpm

Enable CRB repositories

/usr/bin/crb enable

Install Ansible

sudo dnf update
sudo dnf install ansible* -y
ansible-galaxy collection install ansible.posix

Jun 16 '25 12:06 steelhead31

JDK21 tests comparable to the CentOS ones above :

UBI10/x64 (docker):

9 failures in extended.openjdk with 21.0.7+6-LTS and new test material: TAP file, 25 in sanity.openjdk: cd ..TAP file

UBI10/aarch64 (docker):

54 failures in extended.openjdk with GA and latest material: TAP file
Re-run of sanity.openjdk with EA build passed ub about 1h20 (28 passed targets, 0 failed)

RHEL10/x64 (full VM configured with items from DockerStatic dockerfile):

First run with latest material and GA build (so out of sync - can't trivially use old material with RHEL10 due to X11/weston issues) gave 28 failures in sanity.openjdk. TAP file
EA sanity.openjdk passed TAP file
extended.openjdk with last GA and latest material had 71 failures TAP file
extended.openjdk with latest EA: 39 failures

NOTE: Have now also added sw.os.rhel.10 to the labels on the UBI10 machines (to force AQA test to be able to start weston instead of Xvfb) and set XDG_RUNTIME_DIR to /home/jenkins/.xdg-runtime on the machines, plus removed the websocket parameter from the jenkins agent startup line to allow it to work.

Jun 17 '25 09:06 sxa

My run of extended.openjdk with latest nightly jdk21, running on a rhel10 VM (via boxes)

TOTAL: 86 EXECUTED: 43 PASSED: 16 FAILED: 27 DISABLED: 21 SKIPPED: 22 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

TESTCASES RESULTS SUMMARY: passed: 14,332; failed: 89; error: 354; skipped: 0 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Test_openjdk21_hotspot_extended.openjdk_x86-64_linux.tap.txt

Jun 19 '25 07:06 steelhead31

Summary of openjdk AQA runs. I've locked any jobs referenced in this table. These are all JDK21 nightly builds with the latest test material on x64 unless otherwise stated:

Test	OS+virt	Results
sanity	RHEL10 (Boxes)	Pass
sanity	UBI10 (Docker)	1 failure `TestFindINstMemRecursion` (Don't think this recurred)
extended	RHEL10 (Boxes-sxa)	38 failures
extended	RHEL10 (Boxes-SFR)	443 failures (outside jenkins)
extended	UBI10 (Docker)	TBC
extended	UBI9 (Docker)	1 failure `CheckLoopStripMining`
extended	UBI10 (docker-aarch64)	237 failures wayland issues - no failures in nio/net etc.
extended	UBI9 (docker-aarch64)	TBC
extended	CS10-aarch64	All passed
sanity	CS10 (docker)	Based on earlier comment this passed 100%
extended	CS10 (docker)	Based on earlier comment this passed 100%

Limited runs with a subset of the problematic targets (TARGET=testList TESTLIST=jdk_security3_0,jdk_nio_0,jdk_net_0,jdk_rmi_0,jdk_management_0) based on the failures that showed up in the RHEL10 (Boxes-sxa) line above:

OS+virt	Results
RHEL10 (Boxes-sxa)	19 failures
CentOS Stream10 (Docker)	1 failure in `BufferingSubscriberTest`
CentOS Stream 9 (Docker-aarch64)	All passed
CentOS Stream 10 (Docker-aarch64)	All passed

Based on the above, the RHEL10 failures are either some difference between UBI10 and RHEL10 (unlikely) or failures due to the networking in the VM setup that we used for the RHEL10 boxes in lieu of having a cloud provider able to give us RHEL10 at present.

JDK8 results:

Suite	OS+Virt	result
sanity	UBI10 (Docker)	Two failures which I think are down to inconsistent test material re-grind
sanity	RHEL9 (real)	Time Zone failure (inconsistent test material) re-grind failed 10/10
extended	UBI10 (docker)	weston related failures in jdk_beans and jdk_imageio. Algorithm related issues in RMI and security tests
extended	RHEL9 (real)	11 failures (RMI and issues with `libsoftokn3 version not found` which I think mat have a fix somewhere - issue is referenced in https://github.com/adoptium/infrastructure/issues/2900#issuecomment-2488868215)
sanity	RHEL10 (Boxes-sxa) ]	PASSED
extended	RHEL10 (Boxes-sxa)	Some failures including weston related issues - re-running at 242

So JDK8 will show some algorithm related failurs in EL10 (based on the UBI10 results) that are not showing in JDK21.

Jun 19 '25 08:06 sxa

I now have a working RHEL10 VM in virtualbox, Im going to attempt to produce a vagrant box image of this.

Jul 01 '25 12:07 steelhead31

New extended run on RHEL10: https://ci.adoptium.net/job/Grinder/13909/ CONNECTION BROKEN AFTER 15H but seemingly still running. Compared to the RHEL9 run the following suites been executed: jdk_jfr_0 jdk_jfr_1 jdk_instrument_0 jdk_instrument_1 jdk_jdi_0 jdk_jdi_1 jdk_build_0 jdk_build_1 jdk_imageio_0 jdk_imageio_1 jdk_native_sanity_0 jdk_native_sanity_1 jdk_vector_0 jdk_vector_1 with jdk_tools_1 being in progress when it stopped. Restarted all on RHEL10 at https://ci.adoptium.net/job/Grinder/13949 - The process that had hung at that point (processes still there 45 minutes later when other tests are continuing) were:

jenkins  1601207  1.0  2.4 4201736 243952 pts/0  Sl   09:08   0:29 /home/jenkins/workspace/Grinder/jdkbinary/j2sdk-image/bin/java -Xmx512m -jar /home/jenkins/workspace/Grinder/aqa-tests/TKG/../../jvmtest/openjdk/jtreg/lib/jtreg.jar -agentvm -a -ea -esa -v:fail,error,time,nopass -retain:fail,error,*.dmp,javacore.*,heapdump.*,*.trc -ignore:quiet -timeoutFactor:8 -xml:verify -concurrency:3 -nativepath:/home/jenkins/workspace/Grinder/jdkbinary/openjdk-test-image/jdk/jtreg/native -vmoptions:-Xmx512m  -XX:-UseCompressedOops   -w /home/jenkins/workspace/Grinder/aqa-tests/TKG/../TKG/output_1755534508457/jdk_tools_1/work -r /home/jenkins/workspace/Grinder/aqa-tests/TKG/../TKG/output_1755534508457/jdk_tools_1/report -jdk:/home/jenkins/workspace/Grinder/jdkbinary/j2sdk-image -exclude:/home/jenkins/workspace/Grinder/aqa-tests/TKG/../openjdk/openjdk-jdk/test/jdk/ProblemList.txt -exclude:/home/jenkins/workspace/Grinder/aqa-tests/TKG/../../jvmtest/openjdk/excludes/ProblemList_openjdk21.txt -exclude:/home/jenkins/workspace/Grinder/aqa-tests/TKG/../openjdk/excludes/vendors/eclipse/ProblemList_openjdk21.txt /home/jenkins/workspace/Grinder/aqa-tests/TKG/../openjdk/openjdk-jdk/test/jdk:jdk_tools
jenkins  1601400  2.2  2.7 3571584 272900 pts/0  Sl   09:08   1:03 /home/jenkins/workspace/Grinder/jdkbinary/j2sdk-image/bin/java -classpath /home/jenkins/workspace/Grinder/jvmtest/openjdk/jtreg/lib/javatest.jar:/home/jenkins/workspace/Grinder/jvmtest/openjdk/jtreg/lib/jtreg.jar -ea -esa -Xmx512m -XX:-UseCompressedOops --patch-module java.base=/home/jenkins/workspace/Grinder/aqa-tests/TKG/output_1755534508457/jdk_tools_1/work/patches/java.base -Djava.security.policy=file:/home/jenkins/workspace/Grinder/aqa-tests/TKG/../TKG/output_1755534508457/jdk_tools_1/work/jtreg.policy com.sun.javatest.regtest.agent.AgentServer -id 1 -logfile /home/jenkins/workspace/Grinder/aqa-tests/TKG/output_1755534508457/jdk_tools_1/work/jtData/agentServer.1.trace -allowSetSecurityManager -port 53371 -timeoutFactor 8.0
jenkins  1601422  1.7  2.5 3571584 251520 pts/0  Sl   09:08   0:49 /home/jenkins/workspace/Grinder/jdkbinary/j2sdk-image/bin/java -classpath /home/jenkins/workspace/Grinder/jvmtest/openjdk/jtreg/lib/javatest.jar:/home/jenkins/workspace/Grinder/jvmtest/openjdk/jtreg/lib/jtreg.jar -ea -esa -Xmx512m -XX:-UseCompressedOops --patch-module java.base=/home/jenkins/workspace/Grinder/aqa-tests/TKG/output_1755534508457/jdk_tools_1/work/patches/java.base -Djava.security.policy=file:/home/jenkins/workspace/Grinder/aqa-tests/TKG/../TKG/output_1755534508457/jdk_tools_1/work/jtreg.policy com.sun.javatest.regtest.agent.AgentServer -id 2 -logfile /home/jenkins/workspace/Grinder/aqa-tests/TKG/output_1755534508457/jdk_tools_1/work/jtData/agentServer.2.trace -allowSetSecurityManager -port 56087 -timeoutFactor 8.0
jenkins  1601445  1.8  2.4 3571584 240664 pts/0  Sl   09:08   0:53 /home/jenkins/workspace/Grinder/jdkbinary/j2sdk-image/bin/java -classpath /home/jenkins/workspace/Grinder/jvmtest/openjdk/jtreg/lib/javatest.jar:/home/jenkins/workspace/Grinder/jvmtest/openjdk/jtreg/lib/jtreg.jar -ea -esa -Xmx512m -XX:-UseCompressedOops --patch-module java.base=/home/jenkins/workspace/Grinder/aqa-tests/TKG/output_1755534508457/jdk_tools_1/work/patches/java.base -Djava.security.policy=file:/home/jenkins/workspace/Grinder/aqa-tests/TKG/../TKG/output_1755534508457/jdk_tools_1/work/jtreg.policy com.sun.javatest.regtest.agent.AgentServer -id 3 -logfile /home/jenkins/workspace/Grinder/aqa-tests/TKG/output_1755534508457/jdk_tools_1/work/jtData/agentServer.3.trace -allowSetSecurityManager -port 51913 -timeoutFactor 8.0

New extended run on RHEL9: https://ci.adoptium.net/job/Grinder/13910/ PASSED Re-run of failed targets from extended.openjdk 155 from RHEL10 (Boxes-sxa) above (38 failures) testList TESTLIST=jdk_net_0,jdk_net_1,jdk_nio_0,jdk_nio_1,jdk_security3_0,jdk_security3_1,jdk_management_0,jdk_management_1,jdk_rmi_0,jdk_rmi_1 https://ci.adoptium.net/job/Grinder/13911/ Re-run of the RHEL10 38 failing tests on UBI10: https://ci.adoptium.net/job/Grinder/13914/ - One failure in java/net/MulticastSocket/Test. All others passed

50 iterations of the above at G#13914 : All passed other than java/net/MulticastSocket/Test which failed 49/50
another 50 iterations at G#13915: Two failrues of java/net/MulticastSocket/Test
30 iterations on test-sxa-ubi9 at G#13918 - java/net/MulticastSocket/Test failed 1/30 RHEL10/JDK8 extended.openjdk (had some failures but jobs have now expired) https://ci.adoptium.net/job/Grinder/13913/

For the java/net/MulticastSocket/Test I've kicked off some jobs with 1000 iterations. May have been excessive as they're not even half way though after 10 hours but results so far:

test-sxa-ubi9-x64-1: 381 passes so far (Aborted as I needed the machine for other work)
test-docker-ubuntu2204-x64-2: All passed
test-ibmcloud-rhel7-x64-1: All passed
test-docker-fedora42-armv8-1: All passed For the jobs with failures in java/net/MulticastSocket/Test alongside others, I'm running various grinders on UBI10 between 13930 and 13935 with different jobs in the suite to evaluate which one is triggering the failures.

Further evaluation shows that the sequencing seems relevant. Based on running a subset of the tests at G#13490 which caused java/nio/channels/DatagramChannel/AdaptorMulticasting.java to fail 1/2 times after not failing in the runs above I have cut it down further to see when each of the two failing tests work:

G#13945 failed java/net/MulticastSocket/Test after running jck_custom CUSTOM_TARGET=java/net/MulticastSocket/Test.java java/nio/channels/DatagramChannel/MulticastSendReceiveTests.java java/nio/channels/DatagramChannel/AdaptorMulticasting.java java/nio/channels/DatagramChannel/Loopback.java java/nio/channels/spi/SelectorProvider/inheritedChannel/InheritedChannelTest.java - A re-run at G#13957 showed 1/5 failures in each of the AdaptorMulticasting and MulticastSocket/Test tests.
G#13948 PASSED running CUSTOM_TARGET=java/net/MulticastSocket/Test.java java/nio/channels/DatagramChannel/Promiscuous.java java/nio/channels/DatagramChannel/PromiscuousIPv6.java java/nio/channels/DatagramChannel/AdaptorMulticasting.java
Jobs 13951 through 13954 are as 13945 but with one removed off the CUSTOM_TARGET list each time: 13951 ✔️ | 19352 13953 | 13954 | 13955 <<< All passed. Also tried with variations with just three tests and they all passed too 🫢
To re-grind the test in 13945 use this link
Running three tests together in 50 iterations resulted in the two problematic tests failing a small number of times: CUSTOM_TARGET=java/net/MulticastSocket/Test.java java/nio/channels/DatagramChannel/AdaptorMulticasting.java java/nio/channels/DatagramChannel/Loopback.java G#13966

Full re-run of failed targets on RHEL10 gave 38 failures again. custom re-run link running at G#13975 and on UBI10, UBI9, RHEL9, RHEL9 again and SFR's RHEL10

SFR-RHEL10 only failed on the MulticastSocket/Test and CookieHandler B6644726 (See note below) Noting that newer runs, including on RHEL9, are also failing java/net/CookieHandler/B6644726.java. Full e/o on SFR-RHEL10 is in two parts as the first aborted due to a VM issue. Part 1 - Part 2 (Noting that the wayland setup is causing AWT/Swing tests to fail on that machine, but I'm not concerned with that at the moment as those have passed on my Boxes VM)

Aug 18 '25 18:08 sxa

Running other suites on my "Boxes" RHEL10 VM:

Suite/Grinder	result
sanity.perf	✔️
extended.perf	✔️
sanity.system	✔️
extended.system	✔️
sanity.functional	✔️
special.functional	Three timezone test failures in `testList TESTLIST=MBCS_Tests_language_tag_0` [*]
extended.functional	✔️
special.system	✔️
dev.system	✔️
dev.functional	✔️

[*] - failing targets were TWTagTest.timezoneTest, TWTagTest.timezoneTWCalendarTest, TWTagTest.timezoneTWCalendarNumberTest

Aug 20 '25 10:08 sxa

TL;DR Summary of the above:

Most of the test suites pass clean on RHEL10
38 tests in extended.openjdk fail on my "Boxes" VM of RHEL10 - mostly networking/multicast related. These do not seem to occur on Scott's Vagrant/VirtualBox RHEL10 system so the failures should not considered to be blocking and seem to bea result of the VM technology/configuration
If you run a grinder with CUSTOM_TARGET=java/net/MulticastSocket/Test.java java/nio/channels/DatagramChannel/AdaptorMulticasting.java java/nio/channels/DatagramChannel/Loopback.java e.g. G#13968 then you get some intermittent failures (replicated on UBI9 in that Grinder).

Machine	Result of running those three targets 50 times
test-docker-ubuntu2404-x64-1	1/50 failure in adapterMulticasting
test-sxa-rhel10-x64-1	TBC
test-sxa-ubi9-x64-1	Pass but earlier run had 2+3/50 failures
test-rhibmcloud-rhel9-x64-1	Pass
test-docker-fedora42-x64-1	Pass
test-aws-rhel8-x64-1	Pass
test-sxa-ubi10-x64-1	Pass

Aug 21 '25 11:08 sxa

extended.openjdk on other versions

JDK	RHEL10 link	UBI10 suites	UBI10 custom	UBI9 suites	UBI9 custom
21	e.o#155 (38)	-	-	-	-
8	G#14035 SFR (4)	G#14088	G#14056 (4)	G#14055 (3)	G#14057 (3)
11	G#14052 (24)	G#14081 (2)	G#14083 (2)	G#14082 (2)	G#14084 (1)
17	G#14076 (38)	G#14077 (3)	G#14079 (2)	G#14078 (3)	G#14080 (2)

Failures:

CookieHandler B6644726 failed everywhere. Known issue with expired cookie
java/net/MulticastSocket/Test failed about 75% on JDK11, 99% on JDK17 (See earlier analysis) Passed 100X on UBI9
nio DirectIOTest.java.DirectIOTest failed on JDK11 and 17 (although didn't fail at first so wasn't included in custom re-run) - 100 iterations at https://ci.adoptium.net/job/Grinder/14085
One-off failures when re-running full suites:
- nio Wakeup
- net SocketTimeouts
For JDK8, two pkcs11 tests fail (one on UBI9) plus the keytool autotest.sh Re-runs of JDK8 failures (java/net/CookieHandler/B6644726.java sun/security/pkcs11/Provider/Login.sh sun/security/pkcs11/Secmod/AddPrivateKey.java sun/security/tools/keytool/autotest.sh) on other systems:

OS	Result
U2204	Passed
U2404	Passed
F42	Failed All 3
Debian13	Passed
U2504	Passed

Re-runs of JD21 failures (java/net/CookieHandler/B6644726.java java/net/Socket/Timeouts.java java/nio/channels/FileChannel/directio/DirectIOTest.java java/net/MulticastSocket/Test.java) on other systems

OS	Results
UBI10	Consistently fails DirectIOTest (EDIT: Passed on a UBI10 on a host without AV software)
UBI9	Consistently fails DirectIOTest
RHEL8	Passed
RHEL7	Passed
testUBI9	Passed (cf. line 2 issue with host AV? Or btrfs? Nope as it's ok on UBI9 container on my Silverblue btrfs install)
U2404	Passed
F42	Passed

Aug 25 '25 14:08 sxa

Going to try another TL;DR on this one:

On JDK8, Three tests fail on RHEL10 and are believed to be due to the patches for algorithm restrictions on the system NSS which the pkcs11 libraries use: sun/security/pkcs11/Provider/Login.sh, sun/security/pkcs11/Secmod/AddPrivateKey.java, sun/security/tools/keytool/autotest.sh - the second of these is EL10 specific, the other two also fail on EL9. @smlambert as briefly mentioned on a call earlier I feel these should possibly be excluded, even though we can't easily do that just for EL9/10 releases (https://github.com/adoptium/infrastructure/issues/2900 seems a relevant related issue)
nio DirectIOTest.java.DirectIOTest appears to fail if you're running on a machine which has some anti-virus software (identified when running in a container on such a host)
The special.functional suite (JDK21 is the only one tested so far) has three timezone failures in the MBCS_Tests_language_tag_0 target
java/net/MulticastSocket/Test can be variable, and it's possibly related to sequencing of things, and definitely different depending on which JDK version it is, but I don't think that's EL specific. It passed 100 iterations on SFRRHEL10 at java/net/MulticastSocket/Test and again on SFRRHEL10 at G#14150 and on my UBI9 (2) container, but failed 83/100 on my UBI10 container so possibly similar to DirectIOTest ...

Aug 26 '25 16:08 sxa

sanity.openjdk runs

JDK8:

Failed 3 on sxa RHEL10, re-running jdk_util_0 and jdk_util_1 on SFR-RHEL10 also failed UBI10 reruns - 25 iterations - all failed. Fail on "normal" machines too e.g. test-docker-ubuntu2204-x64-2

JDK11:

Failed on sxa RHEL10, re-running langtools_all on SFR-RHEL10 passed The three tests also passed 25 iterations on UBI10 (jdk/shell tests)

JDK17:

Passed on sxa RHEL10

JDK25:

TBC

Also running a full AQA_Test_Pipeline on my UBI10.2 system (Not impacted by AV!) at ~~https://ci.adoptium.net/job/AQA_Test_Pipeline/452/~~ https://ci.adoptium.net/job/AQA_Test_Pipeline/455/ and some more on my N100 system at https://ci.adoptium.net/job/AQA_Test_Pipeline/467/

455 on UBI10 on dockerhost-azure-1had the "known failures" plus one failure in sanity.system (TestJlmRemoteMemoryNoAuth_1: re-run passed) and two IPv6 failures in JD8 extended/openjdk: java/net/ipv6tests/B6521014.java.B6521014 and java/net/Inet6Address/B6206527.java.B6206527 Re-run on 4-ubiwsl was ok. Also running on ubi10-1 (T470p) (Passed), again on dockerhost-azure (Failed), ubi9-1 (T470p) (Passed), Fedora42 (dockerhost-azure (Failed), Alma10 on T420 Ubuntu2204 on dockerhost-azure (Failed), RHEL10 on T470p/boxes (Passed) So it is failing consistently on containers on the dockerhost-azure machine (possibly NAT related?) Link to re-grind Raised separate issue #4046
These are the "known" failures from AQA_Test_Pipeline 455
- JDK11 special.functional, JDK17 special.functional and JDK21 special.functional Three TWTagTest timezone failures
- JDK8 extended.openjdk, JDK11 extended.openjdk, JDK25 extended.openjdk- CookieHandler B6644726 (Note: Passed in JDK17 and JDK21

Aug 27 '25 10:08 sxa

I have created some new machines, of which two are EL10-based:

169.63.181.34 (Centos Stream 10, 2C4G Intel Cascadelake)
161.156.170.114 (Centos Stream 10, 2C8G AMD EPYC)
161.156.80.158 (RHEL8, 2C8G AMD EPYC)
158.176.14.14 (RHEL9, 2C4G Intel Xeon Cascadelake)

Sep 10 '25 10:09 sxa

Playbooks will need some work: fatal: [161.156.80.158]: FAILED! => {"ansible_facts": {}, "changed": false, "failed_modules": {"ansible.legacy.setup": {"ansible_facts": {"discovered_interpreter_python": "/usr/bin/python3"}, "failed": true, "module_stderr": "Shared connection to 161.156.80.158 closed.\r\n", "module_stdout": "/bin/sh: /usr/bin/python3: No such file or directory\r\n", "msg": "The module failed to execute correctly, you probably need to set the interpreter.\nSee stdout/stderr for the exact error", "rc": 127, "warnings": ["No python interpreters found for host 161.156.80.158 (tried ['python3.13', 'python3.12', 'python3.11', 'python3.10', 'python3.9', 'python3.8', '/usr/bin/python3', 'python3'])"]}}, "msg": "The following modules failed to execute: ansible.legacy.setup\n"}

fatal: [161.156.170.114]: FAILED! => {"changed": false, "failures": ["No package epel-release available."], "msg": "Failed to install some of the specified packages", "rc": 1, "results": []}

The latter can be resolved with dnf install -y https://dl.fedoraproject.org/pub/epel/epel-release-latest-10.noarch.rpm although it's "interesting" that this isn't a problem for the static docker images for CS10. SCL install also needs conditional modified for 10.

For the former, RHEL8 seems to have the following list of python packages:

[root@adopttest-rhel8-x64-1-amd ~]# rpm -qa | grep -i python
python3-pyyaml-3.12-12.el8.x86_64
python3-gobject-base-3.28.3-2.el8.x86_64
python3-systemd-234-8.el8.x86_64
python3-cairo-1.16.3-6.el8.x86_64
python3-pydbus-0.6.0-5.el8.noarch
python3-configobj-5.0.6-11.el8.noarch
python3-setools-4.3.0-5.el8.x86_64
python3-setuptools-wheel-39.2.0-9.el8_10.noarch
python3-cloud-what-1.28.44-1.el8_10.x86_64
python3-decorator-4.2.1-2.el8.noarch
python3-libcomps-0.1.18-1.el8.x86_64
python3-pycparser-2.14-14.el8.noarch
python3-oauthlib-2.1.0-1.el8.noarch
python3-jsonpointer-1.10-11.el8.noarch
python3-libsemanage-2.9-12.el8_10.x86_64
python3-netifaces-0.10.6-4.el8.x86_64
python3-magic-5.33-26.el8.noarch
python3-requests-2.20.0-6.el8_10.noarch
python3-rpm-4.14.3-32.el8_10.x86_64
python3-six-1.11.0-8.el8.noarch
python3-libdnf-0.63.0-21.el8_10.x86_64
python3-hawkey-0.63.0-21.el8_10.x86_64
python3-libselinux-2.9-10.el8_10.x86_64
python3-dnf-4.7.0-21.el8_10.noarch
python3-policycoreutils-2.9-26.el8_10.noarch
python3-gobject-3.28.3-2.el8.x86_64
policycoreutils-python-utils-2.9-26.el8_10.noarch
python3-idna-2.5-7.el8_10.noarch
platform-python-setuptools-39.2.0-9.el8_10.noarch
python3-libs-3.6.8-70.el8_10.x86_64
python3-subscription-manager-rhsm-1.28.44-1.el8_10.x86_64
python3-setuptools-39.2.0-9.el8_10.noarch
python3-syspurpose-1.28.44-1.el8_10.x86_64
python3-chardet-3.0.4-7.el8.noarch
python3-ethtool-0.14-5.el8.x86_64
python3-inotify-0.9.6-13.el8.noarch
python3-pip-wheel-9.0.3-24.el8.noarch
platform-python-pip-9.0.3-24.el8.noarch
python3-pyserial-3.1.1-9.el8.noarch
python3-unbound-1.16.2-5.9.el8_10.x86_64
python3-perf-4.18.0-553.70.1.el8_10.x86_64
python3-librepo-1.14.2-5.el8.x86_64
python3-jinja2-2.10.1-7.el8_10.noarch
python3-cffi-1.11.5-6.el8.x86_64
python3-ply-3.9-9.el8.noarch
python3-jwt-1.6.1-2.el8.noarch
python3-pysocks-1.6.8-3.el8.noarch
python3-jsonpatch-1.21-2.el8.noarch
python3-markupsafe-0.23-19.el8.x86_64
python3-audit-3.1.2-1.el8_10.1.x86_64
python3-prettytable-0.7.2-14.el8.noarch
python3-pexpect-4.3.1-3.el8.noarch
python3-babel-2.5.1-7.el8.noarch
python3-dnf-plugins-core-4.0.21-25.el8.noarch
python3-linux-procfs-0.7.3-1.el8.noarch
python3-dateutil-2.6.1-6.el8.noarch
python3-iniparse-0.4-31.el8.noarch
python3-dmidecode-3.12.3-2.el8.x86_64
python3-cryptography-3.2.1-7.el8_9.x86_64
python3-pyudev-0.21.0-7.el8.noarch
python3-dbus-1.2.4-15.el8.x86_64
python3-pytz-2017.2-11.el8.noarch
python3-urllib3-1.24.2-8.el8_10.noarch
platform-python-3.6.8-70.el8_10.x86_64
python3-jsonschema-2.6.0-4.el8.noarch
python3-ptyprocess-0.5.2-4.el8.noarch
python3-gpg-1.13.1-12.el8.x86_64
python3-libxml2-2.9.7-21.el8_10.3.x86_64

So we could override ansible_python_interpreter to be /usr/libexec/platform-python (as dnf uses) or just dnf install python3. For simplicitly and consistency I'm taking the latter approach.

Other things not available in CS10:

i686 packages
ntp
lbzip2

The set default jre operation also had to be commented out for now. Ansible also didn't seem to like RHEL8's python interpreter (3.6.8 in /usr/bin/python3) The current PR at https://github.com/adoptium/infrastructure/pull/4060 allows the CS10 and RHEL9 systems to install but that is very much "in development" and I need to clean that up before it can be merged.

Sep 12 '25 12:09 sxa

Host AQA_pipeline Result

centosstream10-x64-1 (AMD) 480 Failed with the error below

centosstream10-x64-2 (Intel) 484 Failed 8 (jdk_util) and 25 (langtools_all_0) sanity.openjdk - 11/17/21 not yet run. Unresponsive machine - RAM limited? Re-running targets at G#14591 G#14592

rhel9-x64-1 (Intel) 481 Failed 8 and 25 sanity.openjdk. Others ran green/yellow

centosstream10-x64-1 (AMD) 482 All green/yellow inc 25 s/o although 8 s/o didn't run

rhel8-x64-1 (AMD) 483 All yellow/green inc all s/o

Host	AQA_pipeline	Result
centosstream10-x64-1 (AMD)	480	Failed with the error below
centosstream10-x64-2 (Intel)	484	Failed 8 (`jdk_util`) and 25 (`langtools_all_0`) sanity.openjdk - 11/17/21 not yet run. Unresponsive machine - RAM limited? Re-running targets at G#14591 G#14592
rhel9-x64-1 (Intel)	481	Failed 8 and 25 sanity.openjdk. Others ran green/yellow
centosstream10-x64-1 (AMD)	482	All green/yellow inc 25 s/o although 8 s/o didn't run
rhel8-x64-1 (AMD)	483	All yellow/green inc all s/o

All the jobs on centosstream10-x64-2 such as extended.openjdk have failed to clone the material from git, although it seems to have occurred prior to it the pipeline running on the target machine for some reason. Example log

error: RPC failed; curl 92 HTTP/2 stream 5 was not closed cleanly: CANCEL (err 8)
error: 1495 bytes of body are still expected
fetch-pack: unexpected disconnect while reading sideband packet
fatal: early EOF
fatal: fetch-pack: invalid index-pack output

RHEL8 can wait - ~~there is an incompatibility with the current ansible versions and RHEL. Using an older version in a venv is an option. That can be explored separately at some point. I tried installing ansible from the RHEL8 repos to run locally but that will need some work too.~~ EDIT: Second time through I didn't get the same error message about not being able to process a local connection so hopefully that machine is included in the above table now too.

Noting that all versions had an issue with the Set Default JRE/JDK steps in the common role, so that likely needs to be investigated.

Sep 12 '25 17:09 sxa

482 - CS10/AMD

Usualy TWTagTest failures in special.functional. Lots of failures in extended.openjdk due to AWT not being available. One in sanity.openjdk (Also AWT in the hotspot_tier1_compiler_0 target. One extra failure in sanity.openjdk jdk25 in TestLargeUTF8Length. AWT may be related to failed to bind to /tmp/.X11-unix/X0: No such file or directory which I didn't think was fatal but that my be incorrect). Re-grind 17 e/o with jenkins/700 /tmp/.X11-unix created.

481 - RHEL9/Intel

Various issues:

jdk21 e/o LargeGatheringWrite and a couple of runtime ones
jdk21 s/o Two failures in java/fireign TestLargeSegmentCore and one in java_util WhiteBoxResizeTest
jdk25 e/o Failed most of the same as RHEL8/AMD (HttpALot, TestLargeUTF8Length, B6644726 and also a couple of runtime ones.
jdk17 e/o had a couple of failures - one in java/rmi and one runtime/CommandLine

Grinder G#14591 (JDK8, jdk_util_1) (FAILED) and G#14592 (JDK21, langtools_all)

483 - RHEL8/AMD

24 e/o had 14 failures with core dump tests, plus TestLargeUTF8Length as seen in the CS10/AMD sanity.openjdk runs 17 e/o: 6 failures with serviceability/sa tests java.lang.RuntimeException: Test ERROR java.lang.RuntimeException: Unexpected core size: expected 0 > 0 Also some RMI failures (JDK8 particularly) and TestLargeUTF8Length. Plus LargeGathering Write in NIO, CookieHandler B6644726 and a new failure java/net/vthread/HttpALot.java This is a good summary

484 - CS10/Intel

Lots of beans failures (AWT)

Sep 23 '25 12:09 sxa

AQA Pipeline 489 rebuilds 482 (CS10.1) with the .X11-unix fix to aqa-tests. Only s/o and e/o

21 e/o: Failure in java/net/httpclient/http2/H2GoAwayTest.java.H2GoAwayTest
25 e/o: Failures in java/net/CookieHandler/B6644726.java.B6644726, runtime/jni/checked/TestLargeUTF8Length.java.TestLargeUTF8Length and java/nio/channels/Selector/Wakeup.java.Wakeup
25 s/o: Failure in TestLargeUTF8Length
8 e/o. 4 failures: sun/net/www/http/KeepAliveCache/KeepAliveProperty.java.KeepAliveProperty java/rmi/registry/readTest/readTest.sh.readTest java/rmi/registry/reexport/Reexport.java.Reexport sun/rmi/transport/tcp/DeadCachedConnection.java.DeadCachedConnection (Re-grind X1000 at G#14688

AQA Pipeline 490 rebuilds 484 (CS10.2) with the .X11-unix fix to aqa-tests (still running)

Previous pipeline (484) failed 1 in JDK17/e/o: runtime/CommandLine/OptionsValidation/TestOptionsWithRanges.java#id3.TestOptionsWithRanges_id
21 s/o: 2 failures: java/foreign/TestLargeSegmentCopy.java.TestLargeSegmentCopy and java/util/HashMap/WhiteBoxResizeTest.java.WhiteBoxResizeTest
JDK21 e/o: 3 failures: runtime/CommandLine/OptionsValidation/TestOptionsWithRanges.java#id3.TestOptionsWithRanges_id3 java/nio/channels/FileChannel/LargeGatheringWrite.java.LargeGatheringWrite java/rmi/server/Unreferenced/leaseCheckInterval/LeaseCheckInterval.java.LeaseCheckInterval
JDK25 s/o: 4 failures: runtime/jni/checked/TestLargeUTF8Length.java.TestLargeUTF8Length, java/foreign/TestLargeSegmentCopy.java.TestLargeSegmentCopy java/foreign/TestStringEncodingJumbo.java.TestStringEncodingJumbo java/util/HashMap/WhiteBoxResizeTest.java.WhiteBoxResizeTest
JDK11 s/o: 2 failures: ava/util/Currency/CurrencyTest.java.CurrencyTest, java/util/Currency/ValidateISO4217.java.ValidateISO4217
JDK8 e/o 3 failures: sun/net/www/http/KeepAliveCache/KeepAliveProperty.java.KeepAliveProperty, java/rmi/registry/readTest/readTest.sh.readTest, sun/rmi/transport/tcp/DeadCachedConnection.java.DeadCachedConnection
JDK8 s/o java/util/Currency/CurrencyTest.java.CurrencyTest
JDK25 e/o 4 failures: runtime/CommandLine/OptionsValidation/TestOptionsWithRanges.java#id3.TestOptionsWithRanges_id3, runtime/jni/checked/TestLargeUTF8Length.java.TestLargeUTF8Length, java/net/CookieHandler/B6644726.java.B6644726, java/nio/channels/FileChannel/LargeGatheringWrite.java.LargeGatheringWrite

AQA Pipeline 491 rebuilds 483 (RHEL8) with the core fix from Issue 1829 after a successful Grinder

JDK21 e/o failed LargeGatheringWrite - no other failures
JDK8 e/o had two test case failures: java/rmi/registry/readTest/readTest.sh.readTest and sun/rmi/transport/tcp/DeadCachedConnection.java.DeadCachedConnection)
JDK25 e/o had 5 test case failures: runtime/ErrorHandling/CreateCoredumpOnCrash.java.CreateCoredumpOnCrash, java/net/vthread/HttpALot.java.HttpALot, runtime/jni/checked/TestLargeUTF8Length.java.TestLargeUTF8Length, java/net/CookieHandler/B6644726.java.B6644726 and java/nio/channels/FileChannel/LargeGatheringWrite.java.LargeGatheringWrite
JDK11 e/o: 2 failures: java/nio/channels/spi/SelectorProvider/inheritedChannel/InheritedChannelTest.java.InheritedChannelTest and tools/jmod/hashes/HashesOrderTest.java.HashesOrderTest
JDk25 s/o: 2 failures: java/foreign/TestStringEncodingJumbo.java.TestStringEncodingJumbo and runtime/jni/checked/TestLargeUTF8Length.java.TestLargeUTF8Length

AQA Pipeline 494 is openjdk tests on RHEL9 with 512MiB swap enabled and core fix applied

JDK8 sp/o - 3 failures: gc/stress/gclocker/TestExcessGCLockerCollections, java/math/BigInteger/BigIntegerTest, tools/javac/4241573/T4241573
JDK8 e/o - 2 failures: sun/rmi/transport/tcp/DeadCachedConnection.java.DeadCachedConnection, java/rmi/registry/readTest/readTest.sh.readTest
JDK21 e/o: 3 failures - didn't produce tap file as no Test results link? runtime/CommandLine/OptionsValidation/TestOptionsWithRanges.java#id3, java/nio/channels/FileChannel/LargeGatheringWrite.java and java/rmi/server/Unreferenced/leaseCheckInterval/LeaseCheckInterval.java
JDK17 e/o - One hotspot failure: runtime/CommandLine/OptionsValidation/TestOptionsWithRanges.java#id3 (Also didn't show test result link as job listed as failed
JDk25 s/o: 4 failures: java/foreign/TestLargeSegmentCopy.java.TestLargeSegmentCopy java/util/HashMap/WhiteBoxResizeTest.java.WhiteBoxResizeTest ava/foreign/TestStringEncodingJumbo.java.TestStringEncodingJumbo runtime/jni/checked/TestLargeUTF8Length.java.TestLargeUTF8Length
JDK11 s/o: Three failures: runtime/InternalApi/ThreadCpuTimesDeadlock.java.ThreadCpuTimesDeadlock, java/util/Currency/CurrencyTest.java.CurrencyTest java/util/Currency/ValidateISO4217.java.ValidateISO4217
JDK25 e/o failed with a red error status which looks like a connection error at the end, but there were four failures: runtime/CommandLine/OptionsValidation/TestOptionsWithRanges.java#id3, runtime/jni/checked/TestLargeUTF8Length.java, java/net/CookieHandler/B6644726.java, java/nio/channels/FileChannel/LargeGatheringWrite.java
JDK21 s/o had two failures although didn't produce a "Test Results" link: java/util/HashMap/WhiteBoxResizeTest.java java/foreign/TestLargeSegmentCopy.java

AQA Pipeline 498 openjdk on sxa's RHEL10 VM

JDK8 sp/o: 3 failures: gc/stress/gclocker/TestExcessGCLockerCollections, java/math/BigInteger/BigIntegerTest, tools/javac/4241573/T4241573
JDK8 e/o 67 failures, most RMI, some NIO/net - consistent with what we've seen before on the Boxes VM
JDK25 e/o 39 failures, mostly net/nio
JDK8 s/o 1 failure: java/util/Currency/CurrencyTest.java.CurrencyTest

For grinders:

VER	GRINDER	TARGET	CUSTOM_TARGET
8	14662 ✔️	jdk_custom	java/rmi/registry/readTest/readTest.sh sun/rmi/transport/tcp/DeadCachedConnection.java java/math/BigInteger/BigIntegerTest.java
8	14663	hotspot_custom	gc/stress/gclocker/TestExcessGCLockerCollections.java
8	14664 ✔️	langtools_custom	javac/4241573/T4241573.java
25	14665 ⚠️	jdk_custom	java/net/vthread/HttpALot.java java/net/CookieHandler/B6644726.java java/nio/channels/FileChannel/LargeGatheringWrite.java java/nio/channels/Selector/Wakeup.java (X1000 on RHEL8 at G#14689)
25	14666 ✔️	hotspot_custom	runtime/jni/checked/TestLargeUTF8Length.java runtime/ErrorHandling/CreateCoredumpOnCrash.java (X100 on RHEL8 at G#14690)

JDK25's Grinder failed the B6644726 test consistently - all others passed.

Sep 23 '25 15:09 sxa

Of the lots of re-grind jobs from the previous comment:

CS10.1 JDK8 e/o had multiple intermittent failures of differing frequencies (see image below). Trying on three random machines at G#14696 (AWS-RHEL8) ✔️ G#14699 on docker-ubuntu2204 ✔️ G#14700 (Non-docker U2404) ✔️ and on CS10.2 G#14701 and RHEL10 on Boxes G#14702 (worst result - 100% fails on 2/4 tests). Also queued up on on IBMcloud RHEL8 for comparison with the AWS RHEL8 at G#14730 and IBM RHEL9 G#14705 ☣️ (keepalive passed 100%). Grinder on RHEL7 at G#14709✔️ and IBM Ubuntu16.04 ✔️ at G#14710

Also run 100 iterations on CS10.1 at G#14697 ☣️ and 14698 ☣️

ABOVE FAILURES ARE WHEN THE MACHINE'S HOSTNAME RESOLVES TO IPv6 BY DEFAULT. If you force it back by either an entry in /etc/hosts or with -Djava.net.preferIPv4Stack=true then they pass: RHEL9 CS10.1 (although KeepAliveProperty still fails on the CS10 machine) Noting that if you use the external IP address on the CS10 machines in /etc/hosts against the hostname it also fails.

JDK25 on RHEL8 - bit of a mix as per the results shown in the screenshot below. Re-trying on AWS RHEL8 and that seems to be passing (Other than he one CookieHandler one which always fails)

Testing with Forced IPv4 hostname resolution on RHEL9 and RHEL8 - these are not as successful as the JDK8 failures

Sep 27 '25 12:09 sxa

JDK25's LargeGatheringWrite is covered under another issue but some results on these machines since it seems to require special attention on the IBM machines:

ibmcloud-rhel8 PASS
ibmcloud-rhel9 FAIL - Re-run with 4GiB swap enabled
ibmcloud-cs10.1 TBC (Passed in 14766 albeit with "seems to be removed or offline" pauses
ibmcloud-cs10.2 FAIL until 4GiB swap added
aws-rhel8 PASS
sxa-ubi10 PASS
azure-ubi10 PASS

On cs10.2 it is leaving 2GiB files in /tmp). I have confirmed that the jenkins user is able to create larger files (4GiB) successfully so it should not be a ulimit issue.

Test passes on CS10.2 with a 4GiB swapfile (of which it uses over 3GiB) and it creates a file which is 5368709122 bytes in length in /tmp

Sep 30 '25 14:09 sxa

HttpALot is a known issue with a fix being backported at https://bugs.openjdk.org/browse/JDK-8364786

Sep 30 '25 16:09 sxa

Extra runs with all fixes:

JDK25

CS10.1 - JDK25 2 targets Wakeup 2/25 failed
RHEL9 - JDK25 3 tests HttpALot 3/25 failed (Wakeup all passed)
docker U2204 x64-2 JDK25 4 targets Wakeup 4/25 failed
docker U2404 x64-2 JDK25 4 targets Wakeup 1/25 failed
docker UBI9 x64-2 JDK25 4 targets No failures
CS10.1 wakeup test X1000 (Failed on this machine before) Failed 16/1000 iterations
RHEL9 wakeup+LGW All passed
Also CS10.2 Wakeup passing (Failed LGW 15/25 but that was before swap fix)

JDK8

CS10.1 JDK8 KeepAlive failed 30/100
CS10.2 JDK8 Some failures in all four tests - re-run with hostname = localhost in /etc/hosts Failed 74/100 KeepAlive
RHEL8 JDK8 KeepAlive passed, all others had some falures. Re-run with hostname=localhost in /etc/hosts Passed all
RHEL9 JDK8 Passed all
RHEL8 AWS No failures an any of the four tests
UBI10 on laptop (hostname=private IP))
UBI10 on dhx1 (hostname=private IP))
RHEL10 on Boxes (No hostname fix)
CS10 in docker (sxa's laptop) 1 failure out of 100
RHEL10 on Azure Passes
CS10 on Azure Passes

2000 iterations on a few machines based on the fact that it did fail once on a non-IBM cloud system (CS10 on my laptop...)

Azure Ubuntu 24.04
docker CS10 on sxa's laptop 44/2000 failed (6 hr 49 min execution)
IBMcloud CS10.1 (AMD EPYC) 1521/2000 failed (3 hr 21 min execution)
IBMcloud CS10.2 (Intel Xeon) 1506/2000 failed (3 hr 21 min execution)
IBMcloud RHEL9
IBMcloud U1604
docker U2204

Grinders with iterations 1 (to give more time between iterations) are running at Grinders 14842 14843 14844 14845 14846

Subsequent tests with a backported fix in jdk8: RHEL10, UBI10 (sxa's laptop), and TBC CS10.1 TBC CS10.2

Runs on Scott's VirtualBox RHEL10 which are also showing the error similar to the IBMcloud CS10:

First run that you ran without any fix - about 10% failures
Second run with backported fix - passed 100%
Third run with a minimal version of the fix - passed 100%
Fourth run without the fix (to make sure it still fails!) (although slightly fewer failures - about 1/25 so far) So the patch does work.

So CS10 on IBM cloud is failing the KeepAlive tests with a message JavaTest Message: Test threw exception: java.lang.RuntimeException: Failed in server but all others seem ok

(For cut & paste purposes: curl -s https://ci.adoptium.net/job/Grinder/14793/consoleText | grep 'TESTCASES RESULTS SUMMARY')

Sep 30 '25 17:09 sxa

Noting CurrencyTest.java failed on IBMcloud Ubuntu2204 systems - can be re-run with this link - trying that on CS10 (Failed) and RHEL7

Also on three random machines - all failed

G#14819 docker-f42-1
G#14820 ibm-u1604-1
G#14821 azure-u2404-1

EDIT: Passes with ADOPTOPENJDK_BRANCH=v1.0.8-release and the last release build, but not with latest or v1.0.9-release branch. Error is Execution failed: main' threw exception: java.lang.RuntimeException: Test data and JRE's currency data are inconsistent. test: (file: 3 data: 180), JRE: (file: 2 data: 179)` so I think we can ignore this failure for the purposes of this issue's testing.

Sep 30 '25 17:09 sxa

Ansible request for EL10 distributions

CentOS Stream 10 test results

CentOS Stream 9 test results:

CentOS9

CentOS10

CentOS9

CentOS10

UBI10/x64 (docker):

UBI10/aarch64 (docker):

RHEL10/x64 (full VM configured with items from DockerStatic dockerfile):

extended.openjdk on other versions

Failures:

sanity.openjdk runs

JDK8:

JDK11:

JDK17:

JDK25:

482 - CS10/AMD

481 - RHEL9/Intel

483 - RHEL8/AMD

484 - CS10/Intel

AQA Pipeline 489 rebuilds 482 (CS10.1) with the .X11-unix fix to aqa-tests. Only s/o and e/o

AQA Pipeline 490 rebuilds 484 (CS10.2) with the .X11-unix fix to aqa-tests (still running)

AQA Pipeline 491 rebuilds 483 (RHEL8) with the core fix from Issue 1829 after a successful Grinder

AQA Pipeline 494 is openjdk tests on RHEL9 with 512MiB swap enabled and core fix applied

AQA Pipeline 498 openjdk on sxa's RHEL10 VM

JDK25

JDK8