Ceph

Now, 1-2 async op thread can fully meet a OSD's network demand with SSD backend. So maybe we can bind limited thread to special cores, I think it can improve async event loop performance because most of structure and method will processed within thread.

For example,

ms_async_op_threads = 2
ms_async_thread_affinity = 0,3

The size of elements in ms_async_thread_affinity must equal to ms_async_op_threads.