WildFly 40.0.0.Final is released!

[Brian Stansberry]

I'm very happy to announce that the feature-packs for WildFly 40 Final are now available in Maven, with zips and tars at https://www.wildfly.org/downloads/.

You can read all about it in my release announcement post at. https://www.wildfly.org/news/2026/05/21/WildFly-40-is-released/

Some key things:

* The primary WildFly variant now provides EE 11!

* We've introduced a new WildFly EE 10 variant that still provides EE 10. We'll provide that for the 40 and 41 series.

* We've introduced Hashicorp Vault integration via a new feature-pack.

* OIDC logout support.

* Java SE 25 as the recommended SE version for standard WildFly.

... and more.

It has taken a while longer than normal to get this one done, but hopefully you'll think it's worth it.

For WildFly 41 we'll go back to our normal cadence of January/April/July/October releases so we're aiming, so we're aiming to have it out on July 9. You can use the https://www.wildfly.org/events/#calendar page to track our plans.

Best regards,

Brian Stansberry

WildFly project lead

[Markus Schulz]

Hi and thanks,

looks like i've found the first bug? 

14:40:31,919 ERROR [org.jboss.as.controller.management-operation] (Controller Boot Thread) WFLYCTL0013: Operation ("add") failed - address: ([
    ("subsystem" => "jgroups"),
    ("channel" => "ee")
]) - Stage: (VERIFY) - failure description: {"WFLYCTL0080: Failed services" => {"org.wildfly.clustering.jgroups.channel.ee" => "java.util.concurrent.CompletionException: java.lang.IllegalStateException: java.lang.IllegalArgumentException: port out of range:67600
    Caused by: java.util.concurrent.CompletionException: java.lang.IllegalStateException: java.lang.IllegalArgumentException: port out of range:67600
    Caused by: java.lang.IllegalStateException: java.lang.IllegalArgumentException: port out of range:67600
    Caused by: java.lang.IllegalArgumentException: port out of range:67600"}}


Started with port-offset=10000 (-Djboss.socket.binding.port-offset=10000) this offset will now be added to the jgroups-tcp-fd port of 57600.

This was not done for older versions. Is this an intended change?

regards,

Markus

[Markus Schulz]

okay, sorry for the noise.

tcp/FD_SOCK2 seems to be new with Wildfly v40.

[Paul Ferraro]

FD_SOCK2 is not now, but we did resurrect it for the default tcp-based protocol stack for WF40 after encountering too many issues with transport-based failure detection.

You do raise a valid point.  I do not recall if the port-range parameter has any constraints, but it should not allow a value that would push port numbers above the maximum value for an unsigned short.

Additionally, it should also not push port numbers into the ephemeral range that previously were not.

[Paul Ferraro]

/s/now/new

[jpu...@noverant.com]

Hi Brian,

Thanks to you and the team for all the hard work on getting this release out! Very excited to see what we can do with EE 11!

I just have one question: the announcement post mentions that Wildfly 40 now supports virtual threads under Jakarta Concurrency, but when I use `virtual = true` I'm still getting a platform thread back. Am I missing something or is this still the expected behavior?

Regards,

Jonathan Putney

[Brian Stansberry]

I replied to this asking if Jonathan was running on SE 21+ but inadvertently replied to him only and not to the group. He replied saying he was using SE 25.

So this will take some investigation.

[Brian Stansberry]

My apologies, Jonathan, and everyone. I thought the WildFly support for Jakarta Concurrency 3.1's optional virtual thread support feature had landed in WF 40, but I was mistaken. Which was completely my fault. It's still in progress.

I corrected the release announcement -- https://github.com/wildfly/wildfly.org/pull/906/changes / https://www.wildfly.org/news/2026/05/21/WildFly-40-is-released/

Thank you for asking! It's embarrassing  but better to catch it now than confuse even more people!

Best regards,

Brain

[Shalini kuruba]

Hi Brian,

 Unable to build WildFly 40.0.0.Final: wildfly-ejb-client-bom-builder dependency resolution error  

Description

A clean build of WildFly 40.0.0.Final fails during BOM generation.

The failure occurs in:

wildfly-ejb-client-bom-builder

while executing:

org.wildfly.plugins:wildfly-bom-builder-plugin:2.0.10.Final:build-bom

Environment

• WildFly: 40.0.0.Final
• Maven: 3.9.9
• Java: 17
• Linux

Build Command

mvn clean install -Dcheckstyle.skip=true

Error

Failed to execute goal
org.wildfly.plugins:wildfly-bom-builder-plugin:2.0.10.Final:build-bom
on project wildfly-ejb-client-bom-builder

No versions available for
org.osgi.service:org.osgi.service.prefs:jar:[1.1.0,1.2.0)
within specified range

Dependency Chain

org.jboss.hal:hal-console
-> org.jboss.hal:hal-ballroom
-> org.jboss.hal:hal-dmr
-> org.jboss.hal:hal-config
-> org.gwtproject:gwt-dev:2.13.0
-> org.eclipse.jdt:org.eclipse.jdt.core:3.33.0
-> org.eclipse.platform:org.eclipse.core.resources:3.18.200
-> org.eclipse.platform:org.eclipse.core.expressions:3.8.200
-> org.eclipse.platform:org.eclipse.core.runtime:3.12.0
-> org.eclipse.platform:org.eclipse.equinox.preferences:3.10.0
-> org.osgi.service:org.osgi.service.prefs:[1.1.0,1.2.0)

Additional Investigation

The POM for org.eclipse.equinox.preferences:3.10.0 declares:

org.osgi.service org.osgi.service.prefs [1.1.0,1.2.0)

However, Maven Central appears not to contain artifacts under:

org.osgi.service:org.osgi.service.prefs

For example:

https://repo.maven.apache.org/maven2/org/osgi/service/org.osgi.service.prefs/1.1.1/

returns 404.

Can someone confirm whether additional repositories are required or whether this is a known issue with 40.0.0.Final?