Only selectable when specifically requested via "-mca odls pspawn"
Note that there are several concerns:
* we aren't getting SIGCHLD calls when the procs terminate
* we aren't seeing the IO pipes close on termination, though
we are getting output forwarded to mpirun
* I haven't found a way to bind the child process prior to exec.
If we want to use this method, we probably need someone to
implement a cgroup component for the orte/rtc framework
Signed-off-by: Ralph Castain <rhc@open-mpi.org>
Multiple conduits can exist at the same time, and can even point to the same base transport. Each conduit can have its own characteristics (e.g., flow control) based on the info keys provided to the "open_conduit" call. For ease during the transition period, the "legacy" RML interfaces remain as wrappers over the new conduit-based APIs using a default conduit opened during orte_init - this default conduit is tied to the OOB framework so that current behaviors are preserved. Once the transition has been completed, a one-time cleanup will be done to update all RML calls to the new APIs and the "legacy" interfaces will be deleted.
While we are at it: Remove oob/usock component to eliminate the TMPDIR length problem - get all working, including oob_stress
following:
* Provides a fixed number of resource slots (i.e., "hotel rooms").
* Allows one thing to occupy a resource slot at a time (i.e., each
hotel room can have an occupant check in to that room).
* Resource slots can be vacated at any time (i.e., occupants can
voluntarily check out of their hotel room).
* Resource slots can be occupied for a specific maximum amount of
time. If that time expires, the occupant is forcibly evicted and
the upper layer is notified via (libevent) callback (i.e., the maid
will kick an occupant of out of their room when their reservation
is over).
This class can be to be used for things like retransmission schemes
for unreliable transports. For example, a message sent on an
unreliable transport can be checked in to a hotel room. If an ACK for
that message is received, the message can be checked out. But if the
ACK is never received, the message will eventually be evicted from its
room and the upper layer will be notified that the message failed to
check out in time (i.e., that an ACK for that message was not received
in time).
Code using this class is currently being developed off-trunk, but will
be coming to SVN soon.
This commit was SVN r27067.
Roll in the ORTE state machine. Remove last traces of opal_sos. Remove UTK epoch code.
Please see the various emails about the state machine change for details. I'll send something out later with more info on the new arch.
This commit was SVN r26242.
Over the course of time, usage of static ports got corrupted in several places, the "parent" info got incorrectly reset, etc. So correct all that and get the regex-based wireup going again.
Also, don't pass node lists if static ports aren't enabled - they are of no value to the orted and just create the possibility of overly-long cmd lines.
This commit was SVN r24860.
Modify the check_complete code so it finds the first non-zero exit status (i.e., the one from the lowest rank) in a job that terminates normally, and sets the mpirun exit code to that status.
This commit was SVN r23071.
Add orte configuration option to control the use of the framework in the system. Although the code will build, it will not be active unless configured with --enable-bootstrap.
If bootstrap is enabled and the new opal_sysinfo framework can successfully determine the cpu model, pass that info to the application as an MCA param to support some work at Sun.
Also, have daemons report back the resources they find to guide process mapping in bootstrap operations (i.e., where the daemon starts at node boot as opposed to being launched at application start).
Adjust some platform files to enable these capabilities.
This commit was SVN r22244.
Modify the job_complete check so we don't kill jobs when a single proc was terminated by ORTE command via plm.terminate_procs
Still dies gracefully with a ctrl-c, and behaves as before when using plm.terminate_job
This commit was SVN r22227.
1. minor modification to include two new opal MCA params:
(a) opal_profile: outputs what components were selected by each framework
currently enabled for most, but not all, frameworks
(b) opal_profile_file: name of file that contains profile info required
for modex
2. introduction of two new tools:
(a) ompi-probe: MPI process that simply calls MPI_Init/Finalize with
opal_profile set. Also reports back the rml IP address for all
interfaces on the node
(b) ompi-profiler: uses ompi-probe to create the profile_file, also
reports out a summary of what framework components are actually
being used to help with configuration options
3. modification of the grpcomm basic component to utilize the
profile file in place of the modex where possible
4. modification of orterun so it properly sees opal mca params and
handles opal_profile correctly to ensure we don't get its profile
5. similar mod to orted as for orterun
6. addition of new test that calls orte_init followed by calls to
grpcomm.barrier
This is all completely benign unless actively selected. At the moment, it only supports modex-less launch for openib-based systems. Minor mod to the TCP btl would be required to enable it as well, if people are interested. Similarly, anyone interested in enabling other BTL's for modex-less operation should let me know and I'll give you the magic details.
This seems to significantly improve scalability provided the file can be locally located on the nodes. I'm looking at an alternative means of disseminating the info (perhaps in launch message) as an option for removing that constraint.
This commit was SVN r20098.
1. remove direct routed module (hooray!)
2. add radix tree routed module (binomial remains default)
3. remove duplicate data storage - orteds were storing nidmap and pidmap data in odls, everyone else in ess
4. add ess APIs to update nidmap, add new pidmap - used only by orteds for MPI-2 support
5. modify code to eliminate multiple calls to orte_routed.update_route that recreated info already in ess pidmap. Add ess API to lookup that info instead. Modify routed modules to utilize that capability
6. setup new ability to shutdown orteds without sending back an "ack" message to mpirun - not utilized yet, will require some changes to plm terminate_orteds functions in managed environments (coming soon)
Initial tests indicating that fully routing comm via defined routing trees may not actually have a significant cost for operations like IB QP setup. More tests required to confirm.
This will require an autogen...
This commit was SVN r19866.
1. completely and cleanly separates responsibilities between the HNP, orted, and tool components.
2. removes all wireup messaging during launch and shutdown.
3. maintains flow control for stdin to avoid large-scale consumption of memory by orteds when large input files are forwarded. This is done using an xon/xoff protocol.
4. enables specification of stdin recipients on the mpirun cmd line. Allowed options include rank, "all", or "none". Default is rank 0.
5. creates a new MPI_Info key "ompi_stdin_target" that supports the above options for child jobs. Default is "none".
6. adds a new tool "orte-iof" that can connect to a running mpirun and display the output. Cmd line options allow selection of any combination of stdout, stderr, and stddiag. Default is stdout.
7. adds a new mpirun and orte-iof cmd line option "tag-output" that will tag each line of output with process name and stream ident. For example, "[1,0]<stdout>this is output"
This is not intended for the 1.3 release as it is a major change requiring considerable soak time.
This commit was SVN r19767.
Fix race conditions in abnormal terminations. We had done a first-cut at this in a prior commit. However, the window remained partially open due to the fact that the HNP has multiple paths leading to orte_finalize. Most of our frameworks don't care if they are finalized more than once, but one of them does, which meant we segfaulted if orte_finalize got called more than once. Besides, we really shouldn't be doing that anyway.
So we now introduce a set of atomic locks that prevent us from multiply calling abort, attempting to call orte_finalize, etc. My initial tests indicate this is working cleanly, but since it is a race condition issue, more testing will have to be done before we know for sure that this problem has been licked.
Also, some updates relevant to the tool comm library snuck in here. Since those also touched the orted code (as did the prior changes), I didn't want to attempt to separate them out - besides, they are coming in soon anyway. More on them later as that functionality approaches completion.
This commit was SVN r17843.
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.
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.
Clean up the remainder of the size_t references in the runtime itself. Convert to orte_std_cntr_t wherever it makes sense (only avoid those places where the actual memory size is referenced).
Remove the obsolete oob barrier function (we actually obsoleted it a long time ago - just never bothered to clean it up).
I have done my best to go through all the components and catch everything, even if I couldn't test compile them since I wasn't on that type of system. Still, I cannot guarantee that problems won't show up when you test this on specific systems. Usually, these will just show as "warning: comparison between signed and unsigned" notes which are easily fixed (just change a size_t to orte_std_cntr_t).
In some places, people didn't use size_t, but instead used some other variant (e.g., I found several places with uint32_t). I tried to catch all of them, but...
Once we get all the instances caught and fixed, this should once and for all resolve many of the heterogeneity problems.
This commit was SVN r11204.
Note that some compile warnings are generated here because of the direct inclusion of an orte include file in the program. Not entirely sure why that is happening (it is relatively new phenomenon), but it doesn't cause any problems in terms of operation.
This commit was SVN r11175.
Added another system-level test function for ORTE that just spins until terminated by a ctrl-c signal.
Modified orterun - added a couple of newlines to the output when abnormally terminating so the prompt always is on a new line.
This commit was SVN r10866.
1. Modifies the RAS framework so it correctly stores and retrieves the actual slots in use, not just those that were allocated. Although the RAS node structure had storage for the number of slots in use, it turned out that the base function for storing and retrieving that information ignored what was in the field and simply set it equal to the number of slots allocated. This has now been fixed.
2. Modified the RMAPS framework so it updates the registry with the actual number of slots used by the mapping. Note that daemons are still NOT counted in this process as daemons are NOT mapped at this time. This will be fixed in 2.0, but will not be addressed in 1.x.
3. Added a new MCA parameter "rmaps_base_no_oversubscribe" that tells the system not to oversubscribe nodes even if the underlying environment permits it. The default is to oversubscribe if needed and the underlying environment permits it. I'm sure someone may argue "why would a user do that?", but it turns out that (looking ahead to dynamic resource reservations) sometimes users won't know how many nodes or slots they've been given in advance - this just allows them to say "hey, I'd rather not run if I didn't get enough".
4. Reorganizes the RMAPS framework to more easily support multiple components. A lot of the logic in the round_robin mapper was very valuable to any component - this has been moved to the base so others can take advantage of it.
5. Added a new test program "hello_nodename" - just does "hello_world" but also prints out the name of the node it is on.
6. Made the orte_ras_node_t object a full ORTE data type so it can more easily be copied, packed, etc. This proved helpful for the RMAPS code reorganization and might be of use elsewhere too.
This commit was SVN r10697.
1. Changed the RMGR and PLS APIs to add "signal_job" and "signal_proc" entry points. Only the "signal_job" entries are implemented - none of the components have implementations for "signal_proc" at this time. Thus, you can signal all of the procs in a job, but cannot currently signal only one specific proc.
2. Implemented those new API functions in all components except xgrid (Brian will do so very soon). Only the rsh/ssh and fork modules have been tested, however, and only under OS-X.
3. Added signal traps and callback functions for SIGUSR1/2 to orterun/mpirun that catch those signals and call the appropriate commands to propagate them out to all processes in the job.
4. Added a new test directory under the orte branch to (eventually) hold unit and system level tests for just the run-time. Since our test branch of the repository is under restricted access, people working on the RTE were continually developing their own system-level tests - thus making it hard to help diagnose problems. I have moved the more commonly-used functions here, and added one specifically for testing the SIGUSR1/2 functionality.
I will be contacting people directly to seek help with testing the changes on more environments. Other than compile issues, you should see absolutely no change in behavior on any of your systems - this additional functionality is transparent to anyone who does not issue a SIGUSR1/2 to mpirun.
Ralph
This commit was SVN r10258.
