Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.
This project is now read‑only. Starting Monday, February 2, please use https://ibm-ceph.atlassian.net/ for all bug tracking management.

Bug 1553818

Summary: [ceph-ansible][containerized] Container deployment for RGW runs with incorrect role
Product: [Red Hat Storage] Red Hat Ceph Storage Reporter: Tomas Petr <tpetr>
Component: Ceph-AnsibleAssignee: Sébastien Han <shan>
Status: CLOSED DUPLICATE QA Contact: ceph-qe-bugs <ceph-qe-bugs>
Severity: medium Docs Contact:
Priority: medium    
Version: 3.0CC: adeza, aschoen, ceph-eng-bugs, gmeno, mhackett, nthomas, sankarshan, shan
Target Milestone: rc   
Target Release: 3.*   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2018-04-05 10:11:39 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Tomas Petr 2018-03-09 15:42:17 UTC
Description of problem:
The containerized deployment of RadosGW runs daemon inside the container under client.rgw.$(ansible_fqdn), but by the roles/ceph-config/templates/ceph.conf.j2 can deploy ceph.conf only for $(ansible_hostname) == the short hostname.

ceph-ansible is creating ceph.conf for rgw under [tag]
[client.rgw.{{ hostvars[host]['ansible_hostname'] }}]
https://github.com/ceph/ceph-ansible/blob/b1a3c6e4cc87509ade8ab8044fc53c324182c7b2/roles/ceph-config/templates/ceph.conf.j2#L170

where ansible_hostname is "hostname -s"

The parameter like "rgw_use_fqdn" could be introduced as workaround for containerized deployment

Version-Release number of selected component (if applicable):
# rpm -qa | grep ansible
ansible-2.4.2.0-2.el7.noarch
ceph-ansible-3.0.14-1.el7cp.noarch


How reproducible:
always

Steps to Reproduce:
1. deploy ceph containerized environment with RGW
2. add custom values for RGW
3. check the actual configuration of running RGW
 
Actual results:


Expected results:


Additional info:

Comment 4 Sébastien Han 2018-04-05 10:11:39 UTC
We now force the usage of the shortname inside the container so this problem is fixed.
Looks like a dup of: https://bugzilla.redhat.com/show_bug.cgi?id=1546834

The services are different but the issue is the same.
Closing this.

*** This bug has been marked as a duplicate of bug 1546834 ***

Comment 6 Sébastien Han 2018-04-26 11:44:59 UTC
That's a valide concern, the shortname is forced so the socket always has the same name. During the upgrade, the cluster should pick up the new names and update everything accordingly. There is one thing I'm sure aboutt, the monitor directory.