On Wed, Sep 21, 2016 at 4:54 AM, Brian Chang-Chien
> debug rgw = 20
> host = brianceph
> rgw keystone url = http://10.62.13.253:35357
> rgw keystone admin token = 7bb8e26cbc714c47a26ffec3d96f246f
> rgw keystone accepted roles = admin, swiftuser, user, _member_, Member
> rgw keystone token cache size = 500
> rgw keystone revocation interval = 60
> rgw keystone make new tenants = true
> rgw s3 auth use keystone = true
> nss db path = /var/ceph/nss
The debug_rgw=20 has been put into the global section.
I bet that's the sole reason why this particular RadosGW
instance sees it.
> and I still some config problem
> Q3 : when i edit /etc/ceph/ceph.conf , if my hostname is brianceph
> the radosgw term in ceph.conf should be [client.radosgw.gateway] or
> which one is correct?
> PS: when i create radosgw, i call th cmd " ceph-deploy rgw create brianceph"
Most likely your current section naming is wrong. I haven't
poked with ceph-deloy too much but I would say it should
be [client.rgw.brianceph] or [client.radosgw.brianceph].
I don't have now any cluster alive to disambiguate, sorry.
> Q4: when i finish to edit ceph.conf, i need restart radosgw service or
> restart ceph service
> in this case, i use ceph jewel, so which cmd need to call " systemctl
> restart ceph-radosgw.target " or "systemctl ceph.target"
Take a look on that:
> Q5: when i use ceph-deploy new brianceph, ceph will generate a ceph.cof, what
> kind edit ceph.conf to create rgw is prefer
> Method1: i direct edit ceph.conf from ceph geerated, and use ceph-deploy
> --overwrite-conf rgw create brianceph
> Method2(i used in the case) : first i call ceph-deploy rgw create brianceph,
> and then edit ceph.conf in /etc/ceph/ folder , then call systemctl restart
> two methods i find some different issues,
> in Method1, the radosgw item of the ceph.conf in /etc/ceph/ like "rgw
> keystone url" will convert rgw_keystone_url ,
Spaces and underscores in configurables' names are treated
in the same way. No difference.
ceph-users mailing list