Bug 1662131

Document URL: 
1) https://access.redhat.com/documentation/en-us/red_hat_openshift_container_storage/3.11/html-single/deployment_guide/index#idm140254129768144
2) https://access.redhat.com/documentation/en-us/red_hat_gluster_storage/3.3/html-single/container-native_storage_for_openshift_container_platform/#chap-Documentation-Red_Hat_Gluster_Storage_Container_Native_with_OpenShift_Platform-Uninstalling

Section Number and Name: 
For 1) Under Note sections
For 2) Step 3: Cleanup the persistent directories

Describe the issue: 

These links suggest deleting the following directories in order to uninstall OCS from OCP -
~~
# rm -rf /var/lib/heketi \
   /etc/glusterfs \
   /var/lib/glusterd \
   /var/log/glusterfs
~~

When these directories are deleted while the glusterfs package which actually creates these directories is already installed. Now when re-installing OCS the glusterfs package is already installed so it would not get re-installed and these directories do not get created again and then if any playbook which needs to use the /var/log/glusterfs directory, does not get the directory in place and is not able to create a log file which it needs to create in this particular directory and the playbooks fails.

There is already a bug opened for not deleting these directories rather, either deleting the content inside these directories or removing the packages directly which would be again installed when reinstalling OCS.
https://bugzilla.redhat.com/show_bug.cgi?id=1661965

Suggestions for improvement: 
The possible solutions could be to document that either we suggest to remove the packages glusterfs or just to delete the content of there directories rather than removing the directories themselves.
~~
# rm -rf /var/lib/heketi/* \
   /etc/glusterfs/* \
   /var/lib/glusterd/* \
   /var/log/glusterfs/*
~~

Additional information: 
KCS article created for the case -
https://access.redhat.com/solutions/3773501