Global barrier among all worker nodes
For example, assume there are n machines, we want to let machine 0 first
init the values, and then pull the inited value to all machines. Before
pulling, we can place a barrier to guarantee that the initialization is
finished.

 Create a new KVStore
WARNING: it is your responsibility to clear this object through dispose.
- name : #{local, dist} (default is local)
     The type of KVStore
     - local works for multiple devices on a single machine (single process)
     - dist works for multi-machines (multiple processes)

Release the native memory.
The object shall never be used after it is disposed.

Initialize a single or a sequence of key-value pairs into the store.
For each key, one must init it before push and pull.
Only worker 0's (rank == 0) data are used.
This function returns after data have been initialized successfully
kvstore - KVstore
ks - keys (vec or strings or single string)
vs - values (vec or NDArrays or single ndarry)

Load optimizer (updater) state from file
- kvstore
-fname Path to input states file.

Get the number of worker nodes

 Pull a single value or a sequence of values from the store.
 Data consistency:
 1. this function returns after adding an operator to the engine. But any
    further read on out will be blocked until it is finished.
 2. pull is always called after all previous push and pull on the same key are finished
 3. It pulls the newest value from the store.
- kvstore
- ks single or vector of (strings)
- outs single or vector of outs (NDArrays)
- priority
    The priority of the push operation.
    The higher the priority, the faster this action is likely
    to be executed before other push actions.

 Push a single or a sequence of key-value pairs into the store.
 Data consistency:
 1. this function returns after adding an operator to the engine.
 2. push is always called after all previous push and pull on the same key are finished
 3. there is no synchronization between workers. One can use _barrier() to sync all workers

-ks Keys
-vs values  According values
- priority
        The priority of the push operation.
        The higher the priority, the faster this action is likely
        to be executed before other push actions.

Get the rank of this worker node
returns The rank of this node, which is in [0, get_num_workers()) 

Save optimizer (updater) state to file
- kvstore
- fname Path to output states file.

Send a command to all server nodes
 Send a command to all server nodes, which will make each server node run
 KVStoreServer.controller
 This function returns after the command has been executed in all server nodes
-kvstore
-head the head of the command
- body the body of the command

 Whether to do barrier when the kvstore finalizes
- kvstore
- barrier-before-exit boolean

Register an optimizer to the store
If there are multiple machines, this process (should be a worker node)
will pack this optimizer and send it to all servers. It returns after
this action is done

Get the type of the kvstore