manageable_volumes list api timeout with hundreds of rbd volumes in storage pool

[yao ning]

When List summary of volumes available to manage with the cinder API: /v3/{project_id}/manageable_volumes for ceph rbd backends, the restful API call always timeout after 60s.

In the RBD driver, there is neccessary to iterate all rbd volumes to get the watcher status of each rbd volumes. The empty watcher status list tells the volume that is"safe_to_manage" because currently no client connect to the rbd volume.

https://github.com/openstack/cinder/blob/master/cinder/volume/drivers/rbd.py#L1746

any ideas to solve it?