mpi(7) LAM/MPI implementation

DESCRIPTION

LAM features a full implementation of Message Passing Interface (MPI) 1 standard with many features included from the MPI-2 standard.

Compliant applications are source code portable between LAM and any other implementation of MPI. In addition to meeting the standard in a high quality manner, LAM offers extensive monitoring capabilities to support debugging. Monitoring happens on two levels. LAM has the hooks to allow a snapshot of process and message status to be taken at any time during an application run. The status includes all aspects of synchronization plus datatype map / signature, communicator group membership and message contents. On the second level, the MPI library is instrumented to produce a cumulative record of communication, which can be visualized either at runtime or post-mortem.

LAM/MPI features the ability to change its underlying message transport mechanism, as well as pass tunable parameters to different components in LAM/MPI at run-time -- without the need to recompile or relink user MPI applications. This provides great flexibility for both developers of MPI software as well as researchers investigating MPI performance. This modual framework is called the System Services Interface (SSI), and is responsible for much of the back-end functionality in LAM/MPI.

Another strength of this MPI implementation is the movement of non-blocking communication requests, including those that result from buffered sends. This is the real challenge of implementing MPI; everything else is mostly a straight forward wrapping of an underlying communication mechanism. LAM allows messages to be buffered on the source end in any state of progression, including partially transmitted packets. This capability leads to great portability and robustness.

User Information

Users are strongly encouraged to read the LAM/MPI User's Guide that is included with the LAM/MPI distribution, and is provided on the main LAM/MPI web site (http://www.lam-mpi.org/).

Up-to-Date Information

The LAM home page can be found on the World Wide Web at: http://www.lam-mpi.org/. It should be consulted for the most current information about LAM, as well as updates, patches, etc.

MPI Communication

The sophisticated message advancing engine at the heart of the MPI library uses only a handful of routines to drive the underlying communication system. Runtime flags decide which message passing engine module is used, so recompilation of user programs is not necessary.

The different message passing engines are commonly referred to as "Request Progression Interface" (RPI) modules. The LAM/MPI distribution includes multiple RPI modules; see the lamssi_rpi(7) man page for more details.

One notable module uses LAM's network message-passing subsystem, including its buffer daemon. In this "daemon" mode, LAM's extensive monitoring features are fully available. Although the "daemon" mode typically incurrs higher latency than the "native" RPI modules, applications that can utilize latency-hiding techniques may experience greater performance due to the daemon-mode's ability to exhibit true asynchronous message passing.

Guaranteed Envelope Resources

Applications may fail, legitimately, on some implementations but not others due to an escape hatch in the MPI Standard called "resource limitations". Most resources are managed locally and it is easy for an implementation to provide a helpful error code and/or message when a resource is exhausted. Buffer space for message envelopes is often a remote resource (as in LAM) which is difficult to manage. An overflow may not be reported (as in some other implementations) to the process that caused the overflow. Moreover, interpretation of the MPI guarantee on message progress may confuse the debugging of an application that actually died on envelope overflow.

LAM has a property called "Guaranteed Envelope Resources" (GER) which serves two purposes. It is a promise from the implementation to the application that a minimum amount of envelope buffering will be available to each process pair. Secondly, it ensures that the producer of messages that overflows this resource will be throttled or cited with an error as necessary.

A minimum GER is configured when LAM is built. The MPI library uses a protocol to ensure GER when running in daemon mode. The default C2C mode (TCP/IP) does not use a protocol, because process-pair protection is provided by TCP/IP itself. Errors are only reported to the receiving process in C2C mode. An option to mpirun(1) disables GER.

Input and Output

The MPI standard does not specify standard I/O functionality. LAM does not interfere with the I/O capabilities of the underlying system but it does make special provisions for remote terminal I/O using the ANSI/POSIX routines. See mpirun(1) and tstdio(3).

LAM now includes the ROMIO distribution for MPI-2 file input and output. If ROMIO support is compiled into LAM, the functionality from Chapter 9 of the MPI-2 standard is provided.

ROMIO has some important limitations under LAM; the User's Guide in the LAM distribution should be consulted before writing MPI programs that use MPI I/O.

Dynamic Processes

LAM includes an implementation of MPI-2 dynamic process creation.

LAM Extensions to MPI

Debugging Aids

LAM includes the MPI-2 functionality for naming opaque types. Support for the Etnus TotalView parallel debugger is also provided; see the User's Guide for more details.

Additionally, LAM provides the capability to launch non-MPI programs on remote nodes. This includes shell scripts, debuggers, etc. As long as an MPI program is eventually launched (as a child, grandchild, etc.), LAM can handle executing as many intermediate programs as necessary. This can greatly help debugging and logging of user programs.

Trace Generation

To avoid being swamped with trace data from a long running application, LAM supplies collective operations to turn the tap on and off. See MPIL_Trace_on(2) and MPIL_Trace_off(2).

Asynchronous Signals

LAM has an signal handling package which mirrors but does not interfere with POSIX signal handling. An MPI extension routine delivers a signal to a process. See MPIL_Signal(2).

Overview of Commands and Libraries

introu(1), introc(2), INTROF(2)

System Services Interface (SSI)

lamssi(7), lamssi_boot(7), lamssi_coll(7), lamssi_rpi(7)

Starting / Stopping LAM

recon(1), lamboot(1), lamhalt(1), lamnodes(1), lamwipe(1), tping(1), lamgrow(1), lamshrink(1)

Compiling MPI Applications

mpicc(1), mpiCC(1), mpif77(1)

Running MPI Applications

mpirun(1), lamclean(1)

Running Non-MPI Applications

lamexec(1)

Monitoring MPI Applications

mpitask(1)

Unloading MPI Trace Data

lamtrace(1)

Reference Documents

"LAM/MPI Installation Guide"

included in the LAM/MPI distribution and available on http://www.lam-mpi.org/
"LAM/MPI User's Guide"
included in the LAM/MPI distribution and available on http://www.lam-mpi.org/
"LAM Frequently Asked Questions"
at http://www.lam-mpi.org/faq/
"MPI Primer / Developing with LAM", Ohio Supercomputer Center
"MPI: A Message-Passing Interface Standard", Message-Passing Interface Forum, version 1.1
at http://www.mpi-forum.org/
"MPI-2: Extensions to the Message Passing Interface", Message Passing Interface Forum, version 2.0
at http://www.mpi-forum.org/

MPI Quick Tutorials

"LAM/MPI ND User Guide / Introduction"
"MPI: It's Easy to Get Started"
"MPI: Everyday Datatypes"
"MPI: Everyday Collective Communication"
at http://www.lam-mpi.org/mpi/tutorials/lam/

Guaranteed Envelope Resources

"Robust MPI Message Delivery Through Guaranteed Resources",
MPI Developer's Conference, 1995