Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Do not use dockerAgents on ci.adoptium.net #4931

Merged
merged 1 commit into from
Jan 16, 2024
Merged

Conversation

sxa
Copy link
Member

@sxa sxa commented Dec 21, 2023

This is a change to prevent the changes from #4307 from taking effect on the adoptium CI. It can still be used elsewhere, but for now the functionality run these agents does not work on our jenkins instances and causes jobs to be queued up indefinitely waiting for an image which will not be created if all of the real machines are in use. The defined image is also using an older version of qemu which would not be suitable for our needs. We will likely look to resovle these issues going forward.

It may be preferable to have an override at the PLATFORM_MAP level earlier in the file where this is defined, and ultimately we want to have this functionality enabled (For RVV support) but for now this is the easiest way to stop the test failing to run.

Ref: adoptium/temurin-build#3591

@sxa
Copy link
Member Author

sxa commented Dec 21, 2023

Verified in https://ci.adoptium.net/job/Grinder/8327 which queued successfully when given a specific node label to run on

@sxa sxa self-assigned this Dec 21, 2023
@luhenry
Copy link
Contributor

luhenry commented Dec 21, 2023

Do I understand correctly that this is related to "Disable dynamic pool for testing, and queue test jobs for execution on boards"?

@sxa
Copy link
Member Author

sxa commented Dec 21, 2023

Do I understand correctly that this is related to "Disable dynamic pool for testing, and queue test jobs for execution on boards"?

Correct. This should resolve that problem.

@sxa
Copy link
Member Author

sxa commented Dec 22, 2023

Noting that https://ci.adoptium.net/job/Grinder/8349/console seemed to work ok without this change on on dockerhost-equinix-ubuntu2004-x64-1 but that is not universal across our other machines.

dockerAgents = PLATFORM_MAP[params.PLATFORM]["DockerAgents"] ? PLATFORM_MAP[params.PLATFORM]["DockerAgents"] : []
// Block dockerAgents on adoptium CI for riscv64 until we make it work
// This will allow RISC-V test jobs to queue up and not hang
if ( env.JENKINS_URL.contains("ci.adoptium.net")) {
Copy link
Contributor

Choose a reason for hiding this comment

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

This will stop other linux test jobs using dynamic agents. Is this expected?

Copy link
Contributor

Choose a reason for hiding this comment

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

For RISC-V specifically, maybe limit to if ( env.JENKINS_URL.contains("ci.adoptium.net") && SPEC.equals('linux_riscv64')) ?

Copy link
Member Author

@sxa sxa Dec 27, 2023

Choose a reason for hiding this comment

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

I could do, but what other platforms do we do this for? I was under the impression it was only RISC-V and we didn't have the support for it on others

@sxa
Copy link
Member Author

sxa commented Dec 27, 2023

For now I've disabled the machines in the set ci.role.test&&sw.os.linux&&sw.tool.docker&&hw.arch.x86 which are not able to perform this function (verified with the smoke test, so fairly basic) which should allow the test jobs to run through successfully

@smlambert
Copy link
Contributor

Please do not disable https://ci.adoptium.net/label/ci.role.test&&sw.os.linux&&sw.tool.docker&&hw.arch.x86/ machines,
if you do, then external test jobs will fail. We have recently re-enabled them and in 1Q will be working on fixing them to run green. https://ci.adoptium.net/job/Test_openjdk17_hs_sanity.external_x86-64_linux/

The purpose of external tests:
https://github.com/adoptium/aqa-tests/tree/master/external#roadmap

@sxa
Copy link
Member Author

sxa commented Dec 27, 2023

Please do not disable https://ci.adoptium.net/label/ci.role.test&&sw.os.linux&&sw.tool.docker&&hw.arch.x86/ machines, if you do, then external test jobs will fail.

Yep I'm aware of that I've ensured that I'm only changing a subset of them which are being problematic. There should still be enough to run the x64 tests - bear in mind that one of the remaining ones has five executors on it. If you still think that's a problem let me know. This is also an interim measure until this PR can be merged.

@luhenry
Copy link
Contributor

luhenry commented Jan 15, 2024

@sxa @smlambert @sophia-guo how can we move that forward? It's blocking testing for RISC-V on boards-only, which is impossible today for the weekly runs. Happy to have that disabled for RISC-V explicitly. Thanks!

@smlambert
Copy link
Contributor

As a temporary approach, incorporating Sophia's review comments would move this forward.

As the actual approach for this feature, this is not the path I would like us to take.

As per this comment, I would like the images to be an input parameter to the test jobs
#4626 (comment)

Copy link
Contributor

@smlambert smlambert left a comment

Choose a reason for hiding this comment

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

Approving as a temporary approach.

Copy link
Contributor

@karianna karianna left a comment

Choose a reason for hiding this comment

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

Approving since Shelly is OK with it :-)

@sxa sxa merged commit a403063 into adoptium:master Jan 16, 2024
1 check passed
sophia-guo pushed a commit to sophia-guo/openjdk-tests that referenced this pull request Jan 16, 2024
smlambert added a commit that referenced this pull request Jan 16, 2024
* Add missing $(TIMEOUT_HANDLER) to serviceability_jvmti_j9 and others (#4961)

The $(TIMEOUT_HANDLER) is missing from the serviceability_jvmti_j9
playlist. Add it to a number of playlists where it's missing but may be
useful in the future.

Signed-off-by: Peter Shipton <[email protected]>

* aix: exclude java/lang/String/concat/IntegerMinValue.java (#4968)

Signed-off-by: Stewart X Addison <[email protected]>

* Exclude specific Windows testcases (#4965)

* Exclude specific Windows testcases

Signed-off-by: Shelley Lambert <[email protected]>

* Add TestCrashAtIGVNSplitIfSubType

Signed-off-by: Shelley Lambert <[email protected]>

* Exclude JMXStartStopTest on win32 JDK8

Signed-off-by: Shelley Lambert <[email protected]>

* Exclude TestRemoveEmptyCountedLoop on win32

Signed-off-by: Shelley Lambert <[email protected]>

* Remove the jdk8 changes

Signed-off-by: Shelley Lambert <[email protected]>

---------

Signed-off-by: Shelley Lambert <[email protected]>

* Do not use dockerAgents on ci.adoptium.net (#4931)

Signed-off-by: Stewart X Addison <[email protected]>

* Exclude known issues on jdk8 (#4971)

* Exclude known issues on jdk8

Signed-off-by: Sophia Guo <[email protected]>

* Update JMXStartStopTest linked issue

Signed-off-by: Sophia Guo <[email protected]>

---------

Signed-off-by: Sophia Guo <[email protected]>

---------

Signed-off-by: Peter Shipton <[email protected]>
Signed-off-by: Stewart X Addison <[email protected]>
Signed-off-by: Shelley Lambert <[email protected]>
Signed-off-by: Sophia Guo <[email protected]>
Co-authored-by: Peter Shipton <[email protected]>
Co-authored-by: Stewart X Addison <[email protected]>
Co-authored-by: Shelley Lambert <[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