The commit has been tested for C/R and Cray operations, and on Odin (SLURM, rsh) and RoadRunner (TM). I tried to update all environments, but obviously could not test them. I know that Windows needs some work, and have highlighted what is know to be needed in the odls process component.
This represents a lot of work by Brian, Tim P, Josh, and myself, with much advice from Jeff and others. For posterity, I have appended a copy of the email describing the work that was done:
As we have repeatedly noted, the modex operation in MPI_Init is the single greatest consumer of time during startup. To-date, we have executed that operation as an ORTE stage gate that held the process until a startup message containing all required modex (and OOB contact info - see #3 below) info could be sent to it. Each process would send its data to the HNP's registry, which assembled and sent the message when all processes had reported in.
In addition, ORTE had taken responsibility for monitoring process status as it progressed through a series of "stage gates". The process reported its status at each gate, and ORTE would then send a "release" message once all procs had reported in.
The incoming changes revamp these procedures in three ways:
1. eliminating the ORTE stage gate system and cleanly delineating responsibility between the OMPI and ORTE layers for MPI init/finalize. The modex stage gate (STG1) has been replaced by a collective operation in the modex itself that performs an allgather on the required modex info. The allgather is implemented using the orte_grpcomm framework since the BTL's are not active at that point. At the moment, the grpcomm framework only has a "basic" component analogous to OMPI's "basic" coll framework - I would recommend that the MPI team create additional, more advanced components to improve performance of this step.
The other stage gates have been replaced by orte_grpcomm barrier functions. We tried to use MPI barriers instead (since the BTL's are active at that point), but - as we discussed on the telecon - these are not currently true barriers so the job would hang when we fell through while messages were still in process. Note that the grpcomm barrier doesn't actually resolve that problem, but Brian has pointed out that we are unlikely to ever see it violated. Again, you might want to spend a little time on an advanced barrier algorithm as the one in "basic" is very simplistic.
Summarizing this change: ORTE no longer tracks process state nor has direct responsibility for synchronizing jobs. This is now done via collective operations within the MPI layer, albeit using ORTE collective communication services. I -strongly- urge the MPI team to implement advanced collective algorithms to improve the performance of this critical procedure.
2. reducing the volume of data exchanged during modex. Data in the modex consisted of the process name, the name of the node where that process is located (expressed as a string), plus a string representation of all contact info. The nodename was required in order for the modex to determine if the process was local or not - in addition, some people like to have it to print pretty error messages when a connection failed.
The size of this data has been reduced in three ways:
(a) reducing the size of the process name itself. The process name consisted of two 32-bit fields for the jobid and vpid. This is far larger than any current system, or system likely to exist in the near future, can support. Accordingly, the default size of these fields has been reduced to 16-bits, which means you can have 32k procs in each of 32k jobs. Since the daemons must have a vpid, and we require one daemon/node, this also restricts the default configuration to 32k nodes.
To support any future "mega-clusters", a configuration option --enable-jumbo-apps has been added. This option increases the jobid and vpid field sizes to 32-bits. Someday, if necessary, someone can add yet another option to increase them to 64-bits, I suppose.
(b) replacing the string nodename with an integer nodeid. Since we have one daemon/node, the nodeid corresponds to the local daemon's vpid. This replaces an often lengthy string with only 2 (or at most 4) bytes, a substantial reduction.
(c) when the mca param requesting that nodenames be sent to support pretty error messages, a second mca param is now used to request FQDN - otherwise, the domain name is stripped (by default) from the message to save space. If someone wants to combine those into a single param somehow (perhaps with an argument?), they are welcome to do so - I didn't want to alter what people are already using.
While these may seem like small savings, they actually amount to a significant impact when aggregated across the entire modex operation. Since every proc must receive the modex data regardless of the collective used to send it, just reducing the size of the process name removes nearly 400MBytes of communication from a 32k proc job (admittedly, much of this comm may occur in parallel). So it does add up pretty quickly.
3. routing RML messages to reduce connections. The default messaging system remains point-to-point - i.e., each proc opens a socket to every proc it communicates with and sends its messages directly. A new option uses the orteds as routers - i.e., each proc only opens a single socket to its local orted. All messages are sent from the proc to the orted, which forwards the message to the orted on the node where the intended recipient proc is located - that orted then forwards the message to its local proc (the recipient). This greatly reduces the connection storm we have encountered during startup.
It also has the benefit of removing the sharing of every proc's OOB contact with every other proc. The orted routing tables are populated during launch since every orted gets a map of where every proc is being placed. Each proc, therefore, only needs to know the contact info for its local daemon, which is passed in via the environment when the proc is fork/exec'd by the daemon. This alone removes ~50 bytes/process of communication that was in the current STG1 startup message - so for our 32k proc job, this saves us roughly 32k*50 = 1.6MBytes sent to 32k procs = 51GBytes of messaging.
Note that you can use the new routing method by specifying -mca routed tree - if you so desire. This mode will become the default at some point in the future.
There are a few minor additional changes in the commit that I'll just note in passing:
* propagation of command line mca params to the orteds - fixes ticket #1073. See note there for details.
* requiring of "finalize" prior to "exit" for MPI procs - fixes ticket #1144. See note there for details.
* cleanup of some stale header files
This commit was SVN r16364.
2. Remove some unnecessary code that was causing a SEGV.
There may be some more work to be done, but at least orte-clean is functional again.
This commit was SVN r16111.
symbols in them and environ is defined only in the final application
(probably in crt1.o). Apple provides a function for getting at the
environment, so use that instead if it's available.
This commit was SVN r14857.
The primary change that underlies all this is in the OOB. Specifically, the problem in the code until now has been that the OOB attempts to resolve an address when we call the "send" to an unknown recipient. The OOB would then wait forever if that recipient never actually started (and hence, never reported back its OOB contact info). In the case of an orted that failed to start, we would correctly detect that the orted hadn't started, but then we would attempt to order all orteds (including the one that failed to start) to die. This would cause the OOB to "hang" the system.
Unfortunately, revising how the OOB resolves addresses introduced a number of additional problems. Specifically, and most troublesome, was the fact that comm_spawn involved the immediate transmission of the rendezvous point from parent-to-child after the child was spawned. The current code used the OOB address resolution as a "barrier" - basically, the parent would attempt to send the info to the child, and then "hold" there until the child's contact info had arrived (meaning the child had started) and the send could be completed.
Note that this also caused comm_spawn to "hang" the entire system if the child never started... The app-failed-to-start helped improve that behavior - this code provides additional relief.
With this change, the OOB will return an ADDRESSEE_UNKNOWN error if you attempt to send to a recipient whose contact info isn't already in the OOB's hash tables. To resolve comm_spawn issues, we also now force the cross-sharing of connection info between parent and child jobs during spawn.
Finally, to aid in setting triggers to the right values, we introduce the "arith" API for the GPR. This function allows you to atomically change the value in a registry location (either divide, multiply, add, or subtract) by the provided operand. It is equivalent to first fetching the value using a "get", then modifying it, and then putting the result back into the registry via a "put".
This commit was SVN r14711.
This merge adds Checkpoint/Restart support to Open MPI. The initial
frameworks and components support a LAM/MPI-like implementation.
This commit follows the risk assessment presented to the Open MPI core
development group on Feb. 22, 2007.
This commit closes trac:158
More details to follow.
This commit was SVN r14051.
The following SVN revisions from the original message are invalid or
inconsistent and therefore were not cross-referenced:
r13912
The following Trac tickets were found above:
Ticket 158 --> https://svn.open-mpi.org/trac/ompi/ticket/158
remote nodes. It will also kill off rogue orteds and orterun
processes. The killing of processes is ifdef'ed out for Windows
since I do not know how to do it there. Note that this change
will requite an autogen.
This commit was SVN r13477.
memcpy() instead of assigning the struct's by value.
Fixes trac:739.
This commit was SVN r13081.
The following Trac tickets were found above:
Ticket 739 --> https://svn.open-mpi.org/trac/ompi/ticket/739
Sorry for the configure change -- hopefully it's early enough in the
morning that it won't affect people... (new approach won't have a
configure change).
Refs trac:739.
This commit was SVN r13080.
The following Trac tickets were found above:
Ticket 739 --> https://svn.open-mpi.org/trac/ompi/ticket/739
Let's minimize the disturbances and say that the configure system is right.
From now on it's OPAL_BOOL_STRUCT_COPY. This one is related to r13076 and
has to follow when r13076 goes in the 1.2.
This commit was SVN r13077.
The following SVN revision numbers were found above:
r13076 --> open-mpi/ompi@f0932a0701
been fixed in the 7.0 PGI series, but is unlikely to be fixed in the
6.2 series:
* Add a configure test looking for the bad behavior (the PGI compiler
chokes on C code where structs containing bool's are copied by
value)
* Set OMPI_BOOL_STRUCT_COPY to 1 if it's ok, 0 if it's not (i.e., PGI
6.2 series will have this value set to 0)
* In two places in the code base -- orte-clean and btl_openib_ini.h,
we have a struct that contains a bool that is copied by value. In
these two places, check OMPI_BOOL_STRUCT_COPY and if it's 1, use
the "int" type instead of "bool".
Fixes trac:739
This commit was SVN r13076.
The following Trac tickets were found above:
Ticket 739 --> https://svn.open-mpi.org/trac/ompi/ticket/739
but remove them also. This current set of changes will affect
nothing as no one is making use of this ability. However, orte-clean
will be changed soon to utilize this new feature.
This commit was SVN r12996.
- orte-clean.c : check to see if the base session directory is empty
and delete it if it is.
- orte_universe_exists.c : Fix a down stread problem resulting from
George's r10718 commit. Don't use the 'fulldirpath' since
that is no longer guarenteed to be the absolute path
to the session directory. Construct this value outside of that
function from the prefix and frontend vars.
This commit was SVN r10741.
The following SVN revision numbers were found above:
r10718 --> open-mpi/ompi@47eef2e002
After seeing the uglyness that is removing directories in the
codebase I decided to push down this to the OPAL by extending the
opal/os_create_dirpath.(c|h) to contain some more functionality.
In this process I renamed 'os_create_dirpath' to 'os_dirpath' since it
is a bit more general now.
Added a few functions to:
- check if an directory is empty
- check to see if the access permissions are set correctly
- destroy the directory at the end of the dirpath
- By using a caller callback function (a la Perl, I believe)
for every file, the caller can have fine grained control over
whether a specific file is deleted or not.
This simplifies things a bit for orte_session_dir_(finalize|cleanup)
as it should no longer contain any of this functionality, but uses
these functions to do the work.
From the external perspective nothing has changed, from the
developer point of view we have some cleaner, more generic code.
This commit was SVN r10640.
per a request.
Currently it is not working well. That will soon change
as it just needs a bit of attention and testing to
make it lots-mo-betta.
This commit was SVN r10556.
