January 2011 - mpiwg-ft - mailman3.cels.anl.gov

Re: [Mpi3-ft] Run-Through Stabilization Users Guide
by Bronevetsky, Greg 31 Jan '11

31 Jan '11

Josh, the document looks very nice. I'll pass it around LLNL. Some comments: There should be some discussion in the intro about how communication relates to failures (i.e. you can't talk to failed processes but you can talk to non-failed processes). MPI_COMM_VALIDATE is supposed to be used to examine the list of failed ranks in detail after a call to MPI_COMM_VALIDATE_ALL. If there is a failure between the start of MPI_COMM_VALIDATE_ALL and the time MPI_COMM_VALIDATE is called, this list of failed ranks may be larger than the number returned by MPI_COMM_VALIDATE_ALL. There should be a way to remove this inconsistency by allowing the application to carry some kind of connection from MPI_COMM_VALIDATE_ALL to MPI_COMM_VALIDATE to make sure that the application knows which failures have been globally recognized and which only locally. Does collective failure recognition imply local recognition? Would it be better to explicitly have separate states for failures that have been recognized locally and collectively? If nothing else, you should change the language to clearly specify whether local or collective recognition is required. For example, some might get confused and think that getting an error from a collective constitutes recognition or that calling MPI_Comm_validate can make it possible to use collectives after a failure. These constraints are specified earlier in the text but its good to repeat the concept throughout. In the broadcast example it would be easier if both examples used the same bcast algorithm. You don't really explain how the algorithm works, so it'll be easier to understand if you don't switch them. For the Gather example it would be useful to have a picture. Greg Bronevetsky Lawrence Livermore National Lab (925) 424-5756 bronevetsky(a)llnl.gov http://greg.bronevetsky.com > -----Original Message----- > From: mpi3-ft-bounces(a)lists.mpi-forum.org [mailto:mpi3-ft- > bounces(a)lists.mpi-forum.org] On Behalf Of Joshua Hursey > Sent: Friday, January 28, 2011 11:44 AM > To: MPI 3.0 Fault Tolerance and Dynamic Process Control working Group > Subject: [Mpi3-ft] Run-Through Stabilization Users Guide > > As mentioned on the teleconf, I created a Run-Through Stabilization Users > Guide wiki page. > https://svn.mpi-forum.org/trac/mpi-forum- > web/wiki/ft/run_through_users_guide > > The purpose of this page is to introduce application developers to the high > level philosophy of the run-through stabilization proposal, and some of the > most commonly used interfaces and semantics. I included some example code > snippets to help new users start programming against the emerging > prototypes. > > The page is not meant to be comprehensive, but just enough to get an > application developer or new group member started with the ideas contained > in the proposal. This is an alternative to the existing proposal which is > structured to help the group form it into a formal proposal for the MPI > forum, which can be confusing for those not steeped in the the MPI standard > structure and language. > > Let me know what you think, and if you have any suggestions on how to make > the page more useful. > > -- Josh > > ------------------------------------ > Joshua Hursey > Postdoctoral Research Associate > Oak Ridge National Laboratory > http://users.nccs.gov/~jjhursey > > > _______________________________________________ > mpi3-ft mailing list > mpi3-ft(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft

1 0

Re: [Mpi3-ft] Run-Through Stabilization Users Guide
by Joshua Hursey 28 Jan '11

28 Jan '11

George, Feel free to circulate the link to the page. It is meant to help introduce new folks to the developing concepts, and get some early feedback for the group. So the wider the audience the better. Thanks, Josh On Jan 28, 2011, at 3:34 PM, George R Carr Jr wrote: > Hi Josh, > > Hope it is ok for me to forward to a couple others here at NOAA. > -- > George R Carr Jr > U.S. Department of Commerce > NOAA/OAR/GSD > Tel: 303-497-4714 > Email: George.Carr(a)noaa.gov > _______________________________________________ > mpi3-ft mailing list > mpi3-ft(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft > ------------------------------------ Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey

1 0

Re: [Mpi3-ft] Run-Through Stabilization Users Guide
by George R Carr Jr 28 Jan '11

28 Jan '11

Hi Josh, Hope it is ok for me to forward to a couple others here at NOAA. -- George R Carr Jr U.S. Department of Commerce NOAA/OAR/GSD Tel: 303-497-4714 Email: George.Carr(a)noaa.gov

1 0

[Mpi3-ft] Run-Through Stabilization Users Guide
by Joshua Hursey 28 Jan '11

28 Jan '11

As mentioned on the teleconf, I created a Run-Through Stabilization Users Guide wiki page. https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/ft/run_through_users_guide The purpose of this page is to introduce application developers to the high level philosophy of the run-through stabilization proposal, and some of the most commonly used interfaces and semantics. I included some example code snippets to help new users start programming against the emerging prototypes. The page is not meant to be comprehensive, but just enough to get an application developer or new group member started with the ideas contained in the proposal. This is an alternative to the existing proposal which is structured to help the group form it into a formal proposal for the MPI forum, which can be confusing for those not steeped in the the MPI standard structure and language. Let me know what you think, and if you have any suggestions on how to make the page more useful. -- Josh ------------------------------------ Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey

1 0

[Mpi3-ft] FT Working Group Teleconf
by Joshua Hursey 25 Jan '11

25 Jan '11

Reminder that we have a teleconf tomorrow, Wed 1/26/2011, at 12 noon EDT. Date: Jan. 26, 2011 Time: 12 noon EDT/New York Dial-in information: 877-801-8130 Code: 1044056 Agenda: - Piggybacking (Martin S.) https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/Piggybacking - Run-through stabilization proposal (Josh H.) https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/ft/run_through_stabilizat… - other items? Next Meetings: * Feb 7, 2010 (MPI Forum Meeting) * Feb 23, 2010 (Teleconf) * March 9, 2010 (Teleconf) * March 23, 2010 (Teleconf) * March 28, 2010 (MPI Forum Meeting) * April 13, 2010 (Teleconf) * April 27, 2010 (Teleconf) * May 9, 2010 (MPI Forum Meeting) Respond to the list if you have any additional topics that you would like to discuss. -- Josh ------------------------------------ Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey

1 0

[Mpi3-ft] Run-Through Stabilization Proposal Updates (topologies and dynamics)
by Joshua Hursey 11 Jan '11

11 Jan '11

I just pushed in some more updates to the run-through stabilization proposal, change log below: * Added semantics for Ch. 7: Process Topologies * Added note for Ch. 12: External Interfaces * Added semantics for Ch. 10: Process Creation and Management * Added the error classes MPI_ERR_NOT_RESPONDING, MPI_ERR_NOT_AVAILABLE, and MPI_ERR_NOT_CONNECTED for connection establishment routines. * Added the error classes MPI_ERR_FN_CANNOT_CONTINUE, and MPI_ERR_HANDLE_CANNOT_CONTINUE as a way for the MPI library to have more fined grained control over what is disabled due to a previous failure versus the MPI_ERR_CANNOT_CONTINUE error class which is global in scope. * Update Parallel I/O Chapter to include use of the MPI_ERR_HANDLE_CANNOT_CONTINUE, and MPI_ERR_FN_CANNOT_CONTINUE error classes. It is available here: https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/ft/run_through_stabilizat… Other than a few odds and ends in the document, I think the one-sided chapter and more revisions of the parallel I/O chapter are all that is left to do with that document. We'll go over a highlight real of the changes in the teleconf tomorrow if folks want to discuss more there. Enjoy and send comments, Josh ------------------------------------ Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey

1 0

[Mpi3-ft] FT Working Group Teleconf
by Joshua Hursey 11 Jan '11

11 Jan '11

Reminder that we have a teleconf tomorrow, Wed 1/12/2011, at 12 noon EDT. Date: Jan. 12, 2011 Time: 12 noon EDT/New York Dial-in information: 877-801-8130 Code: 1044056 Agenda: - Piggybacking (Greg B., George B., Darius B) https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/Piggybacking - Run-through stabilization proposal (Josh H.) https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/ft/run_through_stabilizat… - Run-through stabilization prototype update / availability - other items? Next Meetings: * Jan 26, 2010 (Teleconf) * Feb 7, 2010 (MPI Forum Meeting) * Feb 23, 2010 (Teleconf) * March 9, 2010 (Teleconf) * March 23, 2010 (Teleconf) * March 28, 2010 (MPI Forum Meeting) * April 13, 2010 (Teleconf) * April 27, 2010 (Teleconf) * May 9, 2010 (MPI Forum Meeting) Respond to the list if you have any additional topics that you would like to discuss. -- Josh ------------------------------------ Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey

1 0

Re: [Mpi3-ft] Teleconf Schedule for 2011: Planning
by George R Carr Jr 06 Jan '11

06 Jan '11

Hi Josh, I may have my first concept running by then. For simplicity, I am using sockets for simplicity. My first test case will be two applications on the same machine so that an exchange of the ip addresses is easily done in the job script. Some rendezvous agent would be needed for future "real" cases but one step at a time. Regards, George P.S. I have copied Tom Henderson (here at NOAA) just to keep him in the loop every so often. >I updated the wiki page with the 2011 teleconf schedule through May, >per an earlier email (at bottom). > >Our next teleconf is next week (Jan. 12). I'll send out call details >and an agenda the day before. Let me know if there is anything in >particular that you want to be discussed and I'll add it to the >agenda. > >-- Josh > >On Dec 7, 2010, at 7:12 PM, Joshua Hursey wrote: > >> We are not scheduled to meet again until 2011 (with holidays and >>travel it's a difficult time to meet before then). So we need to >>decide on a teleconf/meeting schedule for 2011. I propose the >>following to get started, keeping the meeting time at 12/Noon US >>Eastern time and following a rough 2 week spacing. >> >> Jan 12, 2010 (Teleconf) >> Jan 26, 2010 (Teleconf) >> Feb 7, 2010 (MPI Forum Meeting) >> Feb 23, 2010 (Teleconf) >> March 9, 2010 (Teleconf) >> March 23, 2010 (Teleconf) >> March 28, 2010 (MPI Forum Meeting) >> April 13, 2010 (Teleconf) >> April 27, 2010 (Teleconf) >> May 9, 2010 (MPI Forum Meeting) >> >> Let me know if these dates and times work for you. If I don't hear >>anything we will meet again on Jan. 12, 2010 at Noon US Eastern >>time. >> >> -- Josh >> >> ------------------------------------ >> Joshua Hursey >> Postdoctoral Research Associate >> Oak Ridge National Laboratory >> http://users.nccs.gov/~jjhursey >> >> >> _______________________________________________ >> mpi3-ft mailing list >> mpi3-ft(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft >> > >------------------------------------ >Joshua Hursey >Postdoctoral Research Associate >Oak Ridge National Laboratory >http://users.nccs.gov/~jjhursey > > >_______________________________________________ >mpi3-ft mailing list >mpi3-ft(a)lists.mpi-forum.org >http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft -- George R Carr Jr U.S. Department of Commerce NOAA/OAR/GSD Tel: 303-497-4714 Email: George.Carr(a)noaa.gov

1 0

Re: [Mpi3-ft] Teleconf Schedule for 2011: Planning
by Joshua Hursey 06 Jan '11

06 Jan '11

I updated the wiki page with the 2011 teleconf schedule through May, per an earlier email (at bottom). Our next teleconf is next week (Jan. 12). I'll send out call details and an agenda the day before. Let me know if there is anything in particular that you want to be discussed and I'll add it to the agenda. -- Josh On Dec 7, 2010, at 7:12 PM, Joshua Hursey wrote: > We are not scheduled to meet again until 2011 (with holidays and travel it's a difficult time to meet before then). So we need to decide on a teleconf/meeting schedule for 2011. I propose the following to get started, keeping the meeting time at 12/Noon US Eastern time and following a rough 2 week spacing. > > Jan 12, 2010 (Teleconf) > Jan 26, 2010 (Teleconf) > Feb 7, 2010 (MPI Forum Meeting) > Feb 23, 2010 (Teleconf) > March 9, 2010 (Teleconf) > March 23, 2010 (Teleconf) > March 28, 2010 (MPI Forum Meeting) > April 13, 2010 (Teleconf) > April 27, 2010 (Teleconf) > May 9, 2010 (MPI Forum Meeting) > > Let me know if these dates and times work for you. If I don't hear anything we will meet again on Jan. 12, 2010 at Noon US Eastern time. > > -- Josh > > ------------------------------------ > Joshua Hursey > Postdoctoral Research Associate > Oak Ridge National Laboratory > http://users.nccs.gov/~jjhursey > > > _______________________________________________ > mpi3-ft mailing list > mpi3-ft(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft > ------------------------------------ Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey

1 0