Deployment does not complete with OpenShift 3.7.1

[galderz]

I'm having issues with fabric8 maven plugin version 3.5.33 and OpenShift 3.7.1.

When I type mvn fabric8:deploy, the command ends fine but shows this error:

In OpenShift, the deployment does not complete and I see events like this repeated:

11:15:43 AM | app-3-pjzmc | Pod | Warning | Failed | Error: ErrImagePull4 times in the last minute
-- | -- | -- | -- | -- | --
11:15:43 AM | app-3-pjzmc | Pod | Warning | Failed | Failed to pull image "app:latest": rpc error: code = 2 desc = Error response from daemon: {"message":"repository app not found: does not exist or no pull access"}4 times in the last minute
11:15:41 AM | app-3-pjzmc | Pod | Normal | Pulling | pulling image "app:latest"4 times in the last minute
11:15:25 AM | app-3-pjzmc | Pod | Normal | Back-off | Back-off pulling image "app:latest"4 times in the last minute
11:15:25 AM | app-3-pjzmc | Pod | Warning | Failed | Error: ImagePullBackOff4 times in the last minute

Looking at the mvn logs, I see this message appearing:

[INFO] F8: {"kind":"Status","apiVersion":"v1","metadata":{},"status":"Failure","message":"container \"sti-build\" in pod \"app-s2i-2-build\" is waiting to start: PodInitializing","reason":"BadRequest","code":400}

Full logs are here.

Any idea what could be causing this?

Cheers,
Galder

[galderz]

The plugin configuration I'm using is:

        <plugin>
          <groupId>io.fabric8</groupId>
          <artifactId>fabric8-maven-plugin</artifactId>
          <version>${fabric8-maven-plugin.version}</version>
          <executions>
            <execution>
              <id>fmp</id>
              <goals>
                <goal>resource</goal>
                <goal>build</goal>
              </goals>
            </execution>
          </executions>
          <configuration>
            <resources>
              <labels>
                <all>
                  <property>
                    <name>solution</name>
                    <value>${workshop.solution}</value>
                  </property>
                </all>
              </labels>
            </resources>
          </configuration>
        </plugin>

[rohanKanojia]

Hi @galderz ,

Which project are you using for fabric8 maven plugin? Build process in OpenShift 3.7 has become a bit slower as compared to the previous versions. Regarding this container waiting to start error; This issue has been fixed, have a look. This fix is available in 3.5.34, which is now available on maven central. Please see if it solves your problem.

[galderz]

I was trying it out in this project. I'll try 3.5.34.

[galderz]

Same issue happening with 3.5.34, here's the maven log.

[galderz]

This is happening in a different project. I've recreated with verbose Maven logging, here's the more complete output.

[galderz]

Btw, I think there's really a problem here. Here's more info with an example:

OpenShift had version 5 of the application and run mvn fabric8:deploy. Then I call mvn fabric8:deploy which finishes at 13:24:25.

8 minutes later OpenShift eventually deploys version 8 of the application!. You can see here all the events leading up to it.

This is not a problem of being slow, there's a bug and it's simply not usable.

[rohanKanojia]

I think this issue is related to OpenShift 3.7 rather than fabric8 maven plugin, have a look at this issue.

[djohnston89]

same issue on 3.6. works for first deploy, but after undeploying and deploying again we get the same image pull issues.

[rohanKanojia]

@djohnston89 @galderz : Are you guys still facing this issue on Openshift 3.9.0 ? There was a bug in Openshift related to rolling deployments which got back-ported to 3.7.2

[galderz]

I've not used FMP in a while and hence have not tried with recent OpenShift versions

[stale]

This issue has been automatically marked as stale because it has not had any activity since 90 days. It will be closed if no further activity occurs within 7 days. Thank you for your contributions!