Description of problem: I'm trying to create a NFS server in an image. Running from this Dockerfile: ``` FROM fedora:30 MAINTAINER Richard Janík <rjanik> EXPOSE 2049/tcp RUN dnf -y install nfs-utils nc RUN dnf clean all ADD run_nfs /usr/local/bin/run_nfs ENTRYPOINT ["/usr/local/bin/run_nfs"] ``` The `docker build` command on s390x arch shows the following error when installing nfs-utils (the `RUN dnf -y install` line) with: ``` Running scriptlet: nfs-utils-1:2.4.1-0.rc1.fc30.s390x 16/17 System has not been booted with systemd as init system (PID 1). Can't operate. Failed to connect to bus: Host is down ``` The build finishes, but the image is unusable because nfs-utils failed to install correctly. How reproducible: Always reproducible by running docker build with the above Dockerfile. Only reproducible on s390x architecture. Actual results: Build finishes, but nfs-utils were not installed correctly. Expected results: Build finishes, nfs-utils are installed without any issues, the NFS server works. Additional info: Possibly related to a few other issues I found when browsing: https://bugzilla.redhat.com/show_bug.cgi?id=1552976 https://bugzilla.redhat.com/show_bug.cgi?id=1615101
*** This bug has been marked as a duplicate of bug 1615101 ***
I don't think that this is a duplicate of bug 1615101. In that bug, nfs-utils is installed as a dependency of some other packages without any intent of running anything NFS-related, and it causes var-lib-nfs-rpc_pipefs.mount to be added and failing. Here the OP actually tries to run the NFS server.