Description of problem: Scaled registry means having >= 2 registry replicas. Scaled registry requires a storage which provides read-after-write consistency. NFS does not provide read-after-write consistency. Thus the following configuration is invalid: openshift_hosted_registry_replicas=2 openshift_hosted_registry_storage_kind=nfs openshift_hosted_registry_storage_access_modes=['ReadWriteMany'] The installer should not accept and setup this kind of configuration (or any other invalid configuration).
Another example would be ElasticSearch / NFS, see https://docs.openshift.com/container-platform/3.4/install_config/aggregate_logging.html#aggregated-elasticsearch.
PR Created: https://github.com/openshift/openshift-ansible/pull/7867
Backport to 3.9 created: https://github.com/openshift/openshift-ansible/pull/7920
*** Bug 1538298 has been marked as a duplicate of this bug. ***
thanks, Michael! created https://bugzilla.redhat.com/show_bug.cgi?id=1570284 for docs
*** Bug 1538867 has been marked as a duplicate of this bug. ***
Verify this bug with openshift-ansible-3.10.0-0.47.0.git.0.c018c8f.el7.noarch.rpm Set registry or prometheus using nfs storage in ansible inventory, then installer will fail in variable sanity check. openshift_hosted_registry_storage_kind=nfs TASK [Run variable sanity checks] ********************************************** Wednesday 16 May 2018 22:54:02 -0400 (0:00:00.863) 0:00:24.573 ********* fatal: [ec2-54-173-75-179.compute-1.amazonaws.com]: FAILED! => {"failed": true, "msg": "last_checked_host: ec2-54-173-75-179.compute-1.amazonaws.com, last_checked_var: openshift_hosted_registry_storage_kind;nfs is an unsupported type for openshift_hosted_registry_storage_kind. openshift_enable_unsupported_configurations=True mustbe specified to continue with this configuration."} openshift_prometheus_storage_kind=nfs TASK [Run variable sanity checks] ********************************************** Wednesday 16 May 2018 23:18:41 -0400 (0:00:00.828) 0:00:22.202 ********* fatal: [ec2-18-206-191-96.compute-1.amazonaws.com]: FAILED! => {"failed": true, "msg": "last_checked_host: ec2-18-206-191-96.compute-1.amazonaws.com, last_checked_var: openshift_prometheus_storage_kind;nfs is an unsupported type for openshift_prometheus_storage_kind. openshift_enable_unsupported_configurations=True mustbe specified to continue with this configuration."} With openshift_enable_unsupported_configurations=true set in addition, the installation with NFS storage used could continue.
We need to connect https://bugzilla.redhat.com/show_bug.cgi?id=1589137 to this thread.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2018:1816