Hi, I think it is a problem related to 'kubectl' client version. Zeppelin Kubernetes Launcher internally call 'kubectl' command in the container and I guess you can find logs such as
status: ERROR, exception: null, result: %text org.apache.zeppelin.interpreter.InterpreterException: java.io.IOException: error: error validating "STDIN": error validating data: unknown; if you choose to ignore these errors, turn validation off with --validate=false as well. If it is the case, you can change the following lines https://github.com/apache/zeppelin/blob/a15b8f62c3d55197b798c581594a91e9e5220b32/scripts/docker/zeppelin/bin/Dockerfile#L97-L101 to something like ADD https://storage.googleapis.com/kubernetes-release/release/v1.16.3/bin/linux/amd64/kubectl /usr/local/bin/kubectl RUN chmod +x /usr/local/bin/kubectl and install a particular version of Kubectl. Also, please feel free to create a pull-request if necessary. Thanks, moon On Wed, Jan 29, 2020 at 8:54 AM Filippo Balicchia <fbalicc...@gmail.com> wrote: > Hi, > > I'm trying to execute a tiny Spark application in k8s from Zeppelin. In > according with documentation > https://zeppelin.apache.org/docs/0.9.0-SNAPSHOT/quickstart/kubernetes.html, > after build my zeppelin image manually from master `44a49fdf68ecc` deploy > with zeppelin-server.yaml but when I try to run simple code I'm > experiencing the following > org.apache.zeppelin.interpreter.InterpreterException: java.io.IOException: > Error from server (NotFound): pods "spark-enjtgm" not found > > at > org.apache.zeppelin.interpreter.remote.RemoteInterpreter.open(RemoteInterpreter.java:134) > at > org.apache.zeppelin.interpreter.remote.RemoteInterpreter.getFormType(RemoteInterpreter.java:298) > at org.apache.zeppelin.notebook.Paragraph.jobRun(Paragraph.java:424) > at org.apache.zeppelin.notebook.Paragraph.jobRun(Paragraph.java:74) > at org.apache.zeppelin.scheduler.Job.run(Job.java:172) > at > org.apache.zeppelin.scheduler.AbstractScheduler.runJob(AbstractScheduler.java:121) > at > org.apache.zeppelin.scheduler.RemoteScheduler$JobRunner.run(RemoteScheduler.java:159) > at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511) > at java.util.concurrent.FutureTask.run(FutureTask.java:266) > > > It seems that launcher doesn't create a pod then when try to make a spark > action it thrown an exception. > Could you please provide me some tips to understand what I miss in my conf? > > Thanks for help > > --Filippo >
