ceph-radosgw won't start for the other controller nodes in HA

[Lope Beltran II]

I just deployed in HA mode CentOS in 4.1

Given the instruction to manually start ceph-radosgw. After Deployment, I was trying to start the radosgw on 2 of the controllers whose ceph-radosgw is not started. here is the message in the terminal.

Starting radosgw instance(s)...
2014-03-12 02:18:53.273243 7fae36cd9820 -1 WARNING: libcurl doesn't support curl_multi_wait()
2014-03-12 02:18:53.273245 7fae36cd9820 -1 WARNING: cross zone / region transfer performance may be affected
Starting client.radosgw.gateway... [ OK ]

After this, radosgw is not yet running.

I went to the logs and here is the message:

2014-03-12 02:11:59.968833 7f86ada23820 0 ceph version 0.67.5 (a60ac9194718083a4b6a225fc17cad6096c69bd1), process radosgw, pid 21661
2014-03-12 02:11:59.968841 7f86ada23820 -1 WARNING: libcurl doesn't support curl_multi_wait()
2014-03-12 02:11:59.968843 7f86ada23820 -1 WARNING: cross zone / region transfer performance may be affected
2014-03-12 02:11:59.971225 7f86ada23820 -1 asok(0x1fb0600) AdminSocketConfigObs::init: failed: AdminSocket::bind_and_listen: failed to bind the UNIX domain socket to '/var/run/ceph/ceph-client.radosgw.gateway.asok': (13) Permission denied
2014-03-12 02:11:59.985227 7f86ada23820 -1 ERROR: region map does not specify master region
2014-03-12 02:11:59.985732 7f86ada23820 -1 Couldn't init storage provider (RADOS)

[Vladimir Kuklin (vkuklin)]

Would you please attach diagnostic snapshot of the environment?