On 28 August 2015 at 13:05, Nick Coghlan <ncogh...@gmail.com> wrote: > After the other containers also started failing, I tried "setenforce > 0" again with this version of the script, and it looks like the ":Z" > suffix isn't actually giving the different containers the permission > they need to share the host directories. Things might work better with > an exported cross-container volume mount, rather than having a storage > directory on the host.
After reading http://www.infoq.com/articles/docker-executable-images, I decided to attempt the "data container" approach. The script for that is again at https://github.com/ncoghlan/repofunnel/blob/master/_localdev/start_pulp.sh It creates a single "pulp_data" container to manage all the data volumes, and then uses Docker inspect to retrieve the volume details for the two that need remapping to /var/log/httpd for pulpapi and crane (Docker doesn't currently support remapping volume names when using --volume-from). The Python snippets that read the JSON data assume "Mounts" support, so the script requires Docker 1.8+. The other slightly clunky aspect is the fact that Docker doesn't support setting the ownership of mounted volumes to anything other than root, so I had to override the commands for the pulpapi and crane containers to do a chown on /var/lib/pulp. With the current version of the script, that all appears to be working - it's possible to run Pulp locally with that script without touching *any* directories on the host. SELinux still gets very upset if you try to run it in enforcing mode, though. Regards, Nick. -- Nick Coghlan | ncogh...@gmail.com | Brisbane, Australia _______________________________________________ Pulp-list mailing list Pulp-list@redhat.com https://www.redhat.com/mailman/listinfo/pulp-list