Hi Thomas,
On 21/08/2025 15:33, Thomas Wolf wrote:
On 21.08.25 10:59, Emmanuel Lecharny wrote:
Hi,
sorry to bother you, but I have tried to build the project, and it
always ends this way after around 30mins:
[INFO] Running org.apache.sshd.server.shell.InvertedShellWrapperTest
[INFO] Tests run: 3, Failures: 0, Errors: 0, Skipped: 0, Time
elapsed: 0.083 s -- in
org.apache.sshd.server.shell.InvertedShellWrapperTest
[INFO] Running org.apache.sshd.spring.SpringConfigTest
[1]+ Stopped mvn clean install
on my Linux Debian 12 box.
Env:
$ mvn --version
Apache Maven 3.9.9 (8e8579a9e76f7d015ee5ec7bfcdc97d260186937)
Maven home: /opt/maven
Java version: 17.0.16, vendor: Ubuntu, runtime:
/usr/lib/jvm/java-17-openjdk-amd64
Default locale: en_US, platform encoding: UTF-8
OS name: "linux", version: "6.14.0-27-generic", arch: "amd64",
family: "unix"
Of course, I can restart the run (command fg), but this is annoying.
Any idea what would cause such an issue?
No concrete idea. I have never seen this on my Mac, and I don't have
a Linux box.
I never had that on mac either. It happens that I have a linux box now
and running tests on it.
Suspending a running foreground process in a shell is usually Ctrl-Z.
Maybe the shell running the maven process gets a Ctrl-Z somehow?
No. I have launched it many times, on both Java 17 and 21, and I see the
build stop at least 3 times before succeeding (I use fg to restart the
build).
This SpringConfigTest actually runs an interactive OS shell in the
test SSH server. Looking through the code it appears to run
"/bin/sh -i -l" and then an "ls" on Linux.
The test uses JSch to connect to that server. It writes a command,
but actually never consumes the output. It simply disconnects after
one second.
Perhaps any of this results in the shell running maven getting a
Ctrl-Z... I don't see anything in the code (ours or JSch's) that
could cause this, though.
It's rather strange that this didn't happen last time when we did
the 2.15.0 release. I'm not aware of any code changes that might be
related, and while we did bump the JSch version used in tests, I'm
also not aware of any changes in JSch that might be related.
Indeed...
I will relaunch the build in debug mode (mvn -x) to see if I get any
information.
In the 2.15.0 release, you hit a (completely different) build problem
because of the French name for the downloads directory. In what
directory are you running maven this time?
Nope, we are past that:
/home/elecharny/Downloads/apache-sshd-2.16.0
Also, half an hour until these tests get run sounds excessivley long.
The tests are run in sshd-core, twice even, and that happens fairly
early in the whole build. 30 minutes is about the time my old Intel-
based Mac takes for the whole build.
My laptop is pretty powerfull (16 cores, 32Gb or ram).
BTW, I'm not running Debian 12, but Ubuntu 24.04.
In any case, I have voted +1 the release, I assume there is some side
effect on my machine, just wanted to know the root cause of the bizarre
behaviour.
Cheers,
Thomas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@mina.apache.org
For additional commands, e-mail: dev-h...@mina.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@mina.apache.org
For additional commands, e-mail: dev-h...@mina.apache.org
