I reverted rgw nodes to hammer; will wait for jewel 10.2.6 to see if it
fixes the problem and the bug http://tracker.ceph.com/issues/17969 . Thank
you very much for your help.
Regards.
Gagan
On Sat, Mar 4, 2017 at 2:52 PM, K K wrote:
> Hi
>
> I faced in those error. After some
Hi
I faced in those error. After some research I commentded out my custom setting:
#rgw zonegroup root pool = se.root
#rgw zone root pool = se.root
and after those rgw successfully started. Now setting are placed in default
pool: .rgw.root
>Суббота, 4 марта 2017, 6:40 +05:00 от Gagandeep
Permissions are correct see below:
[root@radosgw1 radosgw]# ls -l /etc/ceph
total 16
-rw-r--r--. 1 ceph ceph 63 Nov 19 2015 cephprod.client.admin.keyring
-rw-r--r--. 1 ceph ceph 122 Nov 24 2015 cephprod.client.radosgw.keyring
-rw-r--r--. 1 ceph ceph 1049 Nov 20 2015 cephprod.conf
My first thought is ceph doesn't have permissions to the rados keyring file.
eg.
[root@nuc1 ~]# ls -l /etc/ceph/ceph.client.radosgw.keyring
-rw-rw+ 1 root root 73 Feb 8 20:40
/etc/ceph/ceph.client.radosgw.keyring
You could give it read permission or be clever with setfacl, eg.
setfacl -m
Hi all,
Unable to start radosgw after upgrading hammer(0.94.10) to jewel(10.2.5).
Please see the following log. Can someone help please?
# cat cephprod-client.radosgw.gps-prod-1.log
2017-03-04 10:35:10.459830 7f24316189c0 0 set uid:gid to 167:167
(ceph:ceph)
2017-03-04 10:35:10.459883