Files
lammps/doc/src/message.rst

202 lines
7.2 KiB
ReStructuredText

.. index:: message
message command
===============
Syntax
""""""
.. parsed-literal::
message which protocol mode arg
* which = *client* or *server* or *quit*
* protocol = *md* or *mc*
* mode = *file* or *zmq* or *mpi/one* or *mpi/two*
.. parsed-literal::
*file* arg = filename
filename = file used for message exchanges
*zmq* arg = socket-ID
socket-ID for client = localhost:5555, see description below
socket-ID for server = \*:5555, see description below
*mpi/one* arg = none
*mpi/two* arg = filename
filename = file used to establish communication between 2 MPI jobs
Examples
""""""""
.. code-block:: LAMMPS
message client md file tmp.couple
message server md file tmp.couple
message client md zmq localhost:5555
message server md zmq \*:5555
message client md mpi/one
message server md mpi/one
message client md mpi/two tmp.couple
message server md mpi/two tmp.couple
message quit
Description
"""""""""""
Establish a messaging protocol between LAMMPS and another code for the
purpose of client/server coupling.
The :doc:`Howto client/server <Howto_client_server>` doc page gives an
overview of client/server coupling of LAMMPS with another code where
one code is the "client" and sends request messages to a "server"
code. The server responds to each request with a reply message. This
enables the two codes to work in tandem to perform a simulation.
----------
The *which* argument defines LAMMPS to be the client or the server.
As explained below the *quit* option should be used when LAMMPS is
finished as a client. It sends a message to the server to tell it to
shut down.
----------
The *protocol* argument defines the format and content of messages
that will be exchanged between the two codes. The current options
are:
* md = run dynamics with another code
* mc = perform Monte Carlo moves with another code
For protocol *md*\ , LAMMPS can be either a client or server. See the
:doc:`server md <server_md>` doc page for details on the protocol.
For protocol *mc*\ , LAMMPS can be the server. See the :doc:`server mc <server_mc>` doc page for details on the protocol.
----------
The *mode* argument specifies how messages are exchanged between the
client and server codes. Both codes must use the same mode and use
consistent parameters.
For mode *file*\ , the 2 codes communicate via binary files. They must
use the same filename, which is actually a file prefix. Several files
with that prefix will be created and deleted as a simulation runs.
The filename can include a path. Both codes must be able to access
the path/file in a common filesystem.
For mode *zmq*\ , the 2 codes communicate via a socket on the server
code's machine. Support for socket messaging is provided by the
open-source `ZeroMQ library <http://zeromq.org>`_, which must be
installed on your system. The client specifies an IP address (IPv4
format) or the DNS name of the machine the server code is running on,
followed by a 4 or 5 digit port ID for the socket, separated by a colon.
E.g.
.. parsed-literal::
localhost:5555 # client and server running on same machine
192.168.1.1:5555 # server is 192.168.1.1
deptbox.uni.edu:5555 # server is deptbox.uni.edu
The server specifies "\*:5555" where "\*" represents all available
interfaces on the server's machine, and the port ID must match
what the client specifies.
.. note::
On Linux or Unix machines port IDs below 1024 are reserved to the
superuser and thus not available. Other ports may already be in
use and cannot be opened by a second process. On a Linux machine
the commands "netstat -t4an" or "ss -t4an" will list all locally
used port IDs for IPv4 addresses.
.. note::
On many machines (and sometimes on local networks) also ports IDs
may be blocked by default through firewalls. In that case either
access to the required port (or a desired range of ports) has to
be selectively enabled to the firewall disabled (the latter is
usually not a good idea unless you are on a (small) local network
that is already protected from outside access.
.. note::
Additional explanation is needed here about how to use the *zmq*
mode on a parallel machine, e.g. a cluster with many nodes.
For mode *mpi/one*\ , the 2 codes communicate via MPI and are launched
by the same mpirun command, e.g. with this syntax for OpenMPI:
.. code-block:: bash
mpirun -np 2 lmp_mpi -mpicolor 0 -in in.client -log log.client : -np 4 othercode args # LAMMPS is client
mpirun -np 2 othercode args : -np 4 lmp_mpi -mpicolor 1 -in in.server # LAMMPS is server
Note the use of the "-mpicolor color" command-line argument with
LAMMPS. See the :doc:`command-line args <Run_options>` doc page for
further explanation.
For mode *mpi/two*\ , the 2 codes communicate via MPI, but are launched
be 2 separate mpirun commands. The specified *filename* argument is a
file the 2 MPI processes will use to exchange info so that an MPI
inter-communicator can be established to enable the 2 codes to send
MPI messages to each other. Both codes must be able to access the
path/file in a common filesystem.
----------
Normally, the message client or message server command should be used
at the top of a LAMMPS input script. It performs an initial handshake
with the other code to setup messaging and to verify that both codes
are using the same message protocol and mode. Assuming both codes are
launched at (nearly) the same time, the other code should perform the
same kind of initialization.
If LAMMPS is the client code, it will begin sending messages when a
LAMMPS client command begins its operation. E.g. for the :doc:`fix client/md <fix_client_md>` command, it is when a :doc:`run <run>`
command is executed.
If LAMMPS is the server code, it will begin receiving messages when
the :doc:`server <server>` command is invoked.
If LAMMPS is being used as a client, the message quit command will
terminate its messaging with the server. If you do not use this
command and just allow LAMMPS to exit, then the server will continue
to wait for further messages. This may not be a problem, but if both
the client and server programs were launched in the same batch script,
then if the server runs indefinitely, it may consume the full allocation
of computer time, even if the calculation finishes sooner.
Note that if LAMMPS is the client or server, it will continue
processing the rest of its input script after client/server
communication terminates.
If both codes cooperate in this manner, a new round of client/server
messaging can be initiated after termination by re-using a 2nd message
command in your LAMMPS input script, followed by a new fix client or
server command, followed by another message quit command (if LAMMPS is
the client). As an example, this can be performed in a loop to use a
quantum code as a server to compute quantum forces for multiple LAMMPS
data files or periodic snapshots while running dynamics.
----------
Restrictions
""""""""""""
This command is part of the MESSAGE package. It is only enabled if
LAMMPS was built with that package. See the :doc:`Build package <Build_package>` doc page for more info.
Related commands
""""""""""""""""
:doc:`server <server>`, :doc:`fix client/md <fix_client_md>`
**Default:** none