Bug 1318061

Summary: Rgw commands on RHEL 6.8 are returning ERROR: region root pool name must start with a period
Product: Red Hat Ceph Storage Reporter: Warren <wusui>
Component: RGWAssignee: Ken Dreyer (Red Hat) <kdreyer>
Status: CLOSED NEXTRELEASE QA Contact: ceph-qe-bugs <ceph-qe-bugs>
Severity: medium Docs Contact:
Priority: unspecified    
Version: 1.2.3CC: cbodley, ceph-eng-bugs, gmeno, kbader, mbenjamin, nlevine, owasserm, sweil, wusui, yehuda, yweinste
Target Milestone: rc   
Target Release: 1.3.3   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2016-08-02 20:15:08 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:

Description Warren 2016-03-15 22:28:01 UTC
Description of problem:

sudo radosgw-admin pool list
2016-03-16 13:23:28.248547 7f70ca6bf820 -1 ERROR: region root pool name must start with a period
couldn't init storage provider


Version-Release number of selected component (if applicable):
1.2.3 on RHEL 6.8 

How reproducible:
100% of the time

Steps to Reproduce:
1. Bring up rhel 6.8
2. Install 1.2.3 ceph on it.
3. Attempt to run s3 via rgw.

Note:
In order to test the 6.8 iso, the following steps were performed.							
1. 6.8 was installed from  RHEL-6.8-20160302.0-Server-x86_64-dvd1.iso							
2. The 1.2.3 Ceph iso was downloaded from https://access.redhat.com/downloads  --> Ceph --> Version 1.2 for RHEL 6							
3. ice_setup was installed from the 1.2.3 Ceph iso, and the Ceph-deploy installed by ice_setup							
4. At this point, cdeploy.repo was created in /etc/yum.repos.d with the following content:
							
"[cdeploy]
name = Repo for testing 6.8
baseurl = http://download.eng.rdu2.redhat.com/composes/nightly/RHEL-6.8-20160311.n.0/6/Server/x86_64/os/
gpgcheck = 0
enabled = 1
"						
(look in http://download.eng.rdu.redhat.com/composes/nightly to find the latest version).  The above repo file should be added to all nodes.							
5. Ceph was then installed using ceph-deploy

Comment 1 Warren 2016-03-15 22:31:48 UTC
Note: 6.8 testing was previously tracked by BUG #14982 on Redmine.

Note: This problem may be due to other issues starting radosgw.  I am looking into this.

Comment 2 Warren 2016-03-16 03:28:38 UTC
It appears that the problem here is due the radosgw not starting correctly.

Following the instructions in:

https://access.redhat.com/documentation/en/red-hat-ceph-storage/1.2.3/ceph-object-gateway-for-rhel-x86-64/chapter-2-configure-ceph-object-gateway

In section 2.8, I get:

sudo service ceph-radosgw start
Starting radosgw instance(s)...
/usr/bin/dirname: extra operand `-n'
Try `/usr/bin/dirname --help' for more information.
2016-03-17 02:26:08.359095 7f9b3d8bf820 -1 WARNING: libcurl doesn't support curl_multi_wait()
2016-03-17 02:26:08.359099 7f9b3d8bf820 -1 WARNING: cross zone / region transfer performance may be affected
Starting client.radosgw.gateway...                         [  OK  ]
/usr/bin/radosgw is not running.

Comment 3 Warren 2016-03-16 03:33:32 UTC
Ken.  

The problems here -- I brought up ceph and then used the subscription manager to enable rhel-6-server-rh-common-rpms before installing radosgw.

Is this the right version?  Should I be getting a version from
http://download.eng.rdu2.redhat.com/composes/nightly/RHEL-6.8-20160311.n.0/6/ ?

Comment 4 Warren 2016-03-16 03:41:46 UTC
Ken -- Magna031 currently is the radosgw site that I am trying to install (It's where I ran the service ceph-radosgw start command in comment 2)

Comment 6 Warren 2016-03-18 05:18:20 UTC
Yuri, John: This is part of a smoke test of the 1.2.3 ceph iso on RHEL 6.8.  I don't believe that this affects anything else.

The problems that I encountered earlier were due to the fact that I needed to create a new repo to use the correct libraries for running the ceph-deploy steps.

I suspected that I needed to also find the equivalent libraries for ceph-common and ceph-radosgw.  However, I have not been able to successfully install ceph-radosgw except by getting the rhel-6-server-rh-common-rpms from the subscription manager.
When I use this ceph-radosgw, I get the message '/usr/bin/radosgw is not running.'
whenever I restart httpd.  ps confirms that no radosgw daemon is running and so no
further rados gateway tests run.

So I think that I should be installing something other than the rhel-6-server-rh-common
rpms.  I took a few guesses and tried other branches in http://download.eng.rdu2.redhat.com/composes/nightly/RHEL-6.8-20160311.n.0/6/
but still ran into the same error (/usr/bin/radosgw is not running).

Also, the 6.8 rhel version that I have comes with httpd running.  Httpd start fails with a syntax error on the LoadModule proxy_fcgi_module line that the rgw installation
document claims should be added to /etc/httpd/conf/httpd.conf.  So there may be a
problem with the 6.8 rhel image httpd as well.

Comment 8 Yehuda Sadeh 2016-05-03 18:10:45 UTC
That is fixed in 2.0.

Comment 9 Ken Dreyer (Red Hat) 2016-08-02 20:15:08 UTC
RHCS 1.2 is EOL, and per Comment 8, this will be fixed in RHCS 2.