Open aaronjwhiteside opened 6 years ago
It looks like that the Docker daemon is still not closing the HTTP request to containers/%s/stop
. Its not really clear to me, why this should be the case.
Could you try to add a
<wait>
<kill>2000</kill>
</wait>
to specify a timeout for the stop operation (2s in this example) ?
Setting
<wait>
<kill>2000</kill>
</wait>
made no difference..
Shouldn't DMP have a read timeout? Jenkins killed the job after 2 hours (the integration test timeout value), but I feel like the DMP's read timeout should have kicked in long before that?
Getting the same issue when running a docker image for integration tests and the building a docker image with run skipped.
Setting ```
`` worked the first time and then ran the
mvn clean deploy sonar:sonar` again and it hung again as usual.
Docker Version:
Client:
Version: 18.09.2
API version: 1.39
Go version: go1.10.6
Git commit: 6247962
Built: Sun Feb 10 04:13:47 2019
OS/Arch: linux/amd64
Experimental: false
Server: Docker Engine - Community
Engine:
Version: 18.09.2
API version: 1.39 (minimum version 1.12)
Go version: go1.10.6
Git commit: 6247962
Built: Sun Feb 10 03:42:13 2019
OS/Arch: linux/amd64
Experimental: false
Fabric8 Plugin Version: 0.28.0 Maven version: 3.5.2
Please can this issue be prioritised as it's becoming more and more persistent.
It's not possible to me to reproduce this behaviour. Do you have a trimmed down sample which exhibits this behaviour so that we can troubleshoot this hanging ?
Experienced the issue again with a different project.
Project is a maven multi module with the docker plugin configured in a child pom. The configuration of the plugin in the child pom is pretty much this:
<plugin>
<groupId>io.fabric8</groupId>
<artifactId>docker-maven-plugin</artifactId>
<configuration>
<images>
<image>
<name>artifact</name>
<alias>artifact</alias>
<build>
<dockerFileDir>${project.basedir}/src/main/docker</dockerFileDir>
<assembly>
<inline>
<dependencySets>
<dependencySet>
<includes>
<include>
${project.groupId}:${project.artifactId}
</include>
</includes>
<outputDirectory>.</outputDirectory>
<outputFileNameMapping>application.jar
</outputFileNameMapping>
</dependencySet>
</dependencySets>
</inline>
</assembly>
<tags>
<tag>${project.parent.version}</tag>
<tag>${git.commit.id}</tag>
</tags>
</build>
<run>
<skip>true</skip>
</run>
</image>
<image>
<name>posgres:9.4-alpine
</name>
<alias>db</alias>
<run>
<ports>
<port>5432:5432</port>
</ports>
<wait>
<log>database system is ready to accept connections</log>
<time>10000</time>
</wait>
<log>
<file>${project.build.directory}/db.log</file>
</log>
</run>
</image>
</images>
</configuration>
<executions>
<execution>
<id>start</id>
<phase>pre-integration-test</phase>
<goals>
<goal>start</goal>
</goals>
</execution>
<execution>
<id>stop</id>
<phase>post-integration-test</phase>
<goals>
<goal>stop</goal>
</goals>
</execution>
<execution>
<id>docker-build</id>
<goals>
<goal>build</goal>
</goals>
</execution>
<execution>
<id>docker-push</id>
<goals>
<goal>push</goal>
</goals>
</execution>
</executions>
</plugin>
With the above configuration I downgraded to version 0.27.2 of the docker plugin and the issue occurs there to.
@rhuss any movement on this, given @emmaLP 's example?
Sorry, no. Still stuck in work and not sure when I can pick this up again.
Same problem as #552?
Description
On our jenkins server/slave about half the time the dmp fails to stop the containers it started.. on closer inspection it appears that the container did exit, but that dmp failed to detect this..
jstack on the maven process
Info
d-m-p version : 0.25.2
Maven version (
mvn -v
) :Docker version :
Server: Engine: Version: 17.12.1-ce API version: 1.35 (minimum version 1.12) Go version: go1.9.4 Git commit: 7390fc6/17.12.1-ce Built: Tue Apr 3 23:38:52 2018 OS/Arch: linux/amd64 Experimental: false