May 2012 - mpiwg-ft - mailman3.cels.anl.gov

Re: [Mpi3-ft] Updated FT Chapter
by Supalov, Alexander 10 May '12

10 May '12

Hi Bland, Thanks. It appears that table 17.1 between pages 545 and 546 ran away from Section 17.4 where there's a colon awaiting it. LaTeX, I suppose. Best regards. Alexander From: mpi3-ft-bounces(a)lists.mpi-forum.org [mailto:[email protected]] On Behalf Of Wesley Bland Sent: Wednesday, May 09, 2012 3:20 PM To: mpi3-ft(a)lists.mpi-forum.org Subject: Re: [Mpi3-ft] Updated FT Chapter Clearly the number of changes involved in renaming AGREEMENT to AGREE and INVALIDATE to REVOKE was overwhelming and made the document difficult to read. I've removed all of those changes and attached another draft which should be much simpler to parse. Wesley On Tue, May 8, 2012 at 3:56 PM, Wesley Bland <wbland(a)eecs.utk.edu<mailto:[email protected]>> wrote: We've completed the new draft of the FT chapter. I've attached just the FT chapter to this email. The rest of the document can be found on the ticket (https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323) In this new version, we've clarified many of the points raised at the forum meetings and on the recent conference calls. Here's a list: - MPI_COMM_INVALIDATE is now MPI_COMM_REVOKE due to naming conflicts (along with the RMA and I/O versions) - MPI_COMM_AGREEMENT is now MPI_COMM_AGREE to follow naming conventions - A paragraph was added to 17.1 (Introduction) to clarify that FT is optional, but implementing the new FT functions is not. - Section 17.2.2 (Point-to-Point and Collective Communication) has been reworked to try to make it more clean when to return MPI_ERR_PENDING and MPI_ERR_PROC_FAILED. - The non-rooted collective discussion in the advice to users in Section 17.2.2 has been changed to make it more clear when collective operations might have non-uniform return values. - The definition of MPI_COMM_REVOKE has been clarified to further define that the function is not collective and does not have a matching call on remote processes. If you see more changes that you think would be helpful, please let us know either on this list or on the conference call tomorrow. Thanks, Wesley Bland -------------------------------------------------------------------------------------- Intel GmbH Dornacher Strasse 1 85622 Feldkirchen/Muenchen, Deutschland Sitz der Gesellschaft: Feldkirchen bei Muenchen Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer Registergericht: Muenchen HRB 47456 Ust.-IdNr./VAT Registration No.: DE129385895 Citibank Frankfurt a.M. (BLZ 502 109 00) 600119052

1 0

Re: [Mpi3-ft] Updated FT Chapter
by Wesley Bland 09 May '12

09 May '12

Clearly the number of changes involved in renaming AGREEMENT to AGREE and INVALIDATE to REVOKE was overwhelming and made the document difficult to read. I've removed all of those changes and attached another draft which should be much simpler to parse. Wesley On Tue, May 8, 2012 at 3:56 PM, Wesley Bland <wbland(a)eecs.utk.edu> wrote: > We've completed the new draft of the FT chapter. I've attached just the FT > chapter to this email. The rest of the document can be found on the ticket ( > https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323) > > In this new version, we've clarified many of the points raised at the > forum meetings and on the recent conference calls. Here's a list: > > - MPI_COMM_INVALIDATE is now MPI_COMM_REVOKE due to naming conflicts > (along with the RMA and I/O versions) > - MPI_COMM_AGREEMENT is now MPI_COMM_AGREE to follow naming conventions > - A paragraph was added to 17.1 (Introduction) to clarify that FT is > optional, but implementing the new FT functions is not. > - Section 17.2.2 (Point-to-Point and Collective Communication) has been > reworked to try to make it more clean when to return MPI_ERR_PENDING and > MPI_ERR_PROC_FAILED. > - The non-rooted collective discussion in the advice to users in Section > 17.2.2 has been changed to make it more clear when collective operations > might have non-uniform return values. > - The definition of MPI_COMM_REVOKE has been clarified to further define > that the function is not collective and does not have a matching call on > remote processes. > > If you see more changes that you think would be helpful, please let us > know either on this list or on the conference call tomorrow. > > Thanks, > Wesley Bland >

1 0

[Mpi3-ft] FT WG Teleconference Wednesday @ Noon EDT
by Josh Hursey 09 May '12

09 May '12

Reminder that we have a teleconf Wednesday at Noon EDT. Date: May 9, 2012 Time: Noon EDT/New York Dial-in information: 877-801-8130 Code: 1044056 Agenda: - Teleconf schedule - extended - User-Level Failure Mitigation Proposal https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323 - Timeline to vote: - 5/7: Finish ticket 0 edits to text (UTK) - 5/14: Include new fortran interfaces in text (ORNL) - 5/14: Start outlining slides for MPI Forum (George/all) - 5/14: Circulate final ticket 0 edited text to MPI Forum (all) - 5/21: Make implementation available (UTK/ORNL) - 5/21: Finish draft of slides for MPI Forum (George/all) - 5/28: MPI Forum meeting (Vote 5/30) - Touch base on new draft http://lists.mpi-forum.org/mpi3-ft/2012/05/1125.php - Implementation progress - Dynamics sub-ticket https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/327 - I/O sub-ticket https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/326 - RMA sub-ticket https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/325 - Other items? Additional Agenda Items [as time permits]: - FT Query Interface (Josh) https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/336 - FATAL scope clarification (Josh) https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/324 - FA-MPI Alternative Proposal (Tony) http://lists.mpi-forum.org/mpi3-ft/2012/02/1048.php - Piggybacking (Martin S.) https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/Piggybacking - Process Recovery (Josh H.) https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/ft/process_recovery_2 - Error to Fault changes proposal (Darius B.) http://lists.mpi-forum.org/mpi3-ft/2011/06/0773.php - Network Channel Failure http://lists.mpi-forum.org/mpi3-ft/2011/06/0730.php Next Meetings: * May 16, 2012 (Teleconf - If needed) * May 23, 2012 (Teleconf) * May 28, 2012 (MPI Forum Meeting) * June 6, 2012 (Teleconf) * June 13, 2012 (Teleconf - If needed) * June 20, 2012 (Teleconf) * June 27, 2012 (Teleconf - If needed) * July 4, 2012 (No meeting - US Holiday) * July 11, 2012 (Teleconf) * July 16, 2012 (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] Updated FT Chapter
by Bronis R. de Supinski 08 May '12

08 May '12

Re: > what is exactly meant with "in which all processes contribute to the > result and all processes receive the result" - what is "the result"? > This doesn't really describe an alltoall, since any process only > receives part of a "result". Perhaps this should be reformulated to "in > which all processes contribute data and all processes receive data"? > This probably doesn't catch MPI_Scan, though. Or, perhaps it should use the existing terminology in Section 5.2.2 on pages 134-135 of the 2.2 document? This issue is one of the many problems with the proposed changes. They are not consistent with the existing document. It is one of the many examples of how the changes were assembled in a hurried fashion that is not appropriate for something that would become a lasting portion of the standard if accepted. The appropriate thing is for those who proposed this ticket to withdraw it although I harbor no illusions that such actions will occur. Bronis

1 0

Re: [Mpi3-ft] Updated FT Chapter
by Schulz, Martin 08 May '12

08 May '12

Hi all, Unfortunately, I won't make it tomorrow due to another conflicting call, but I wanted to raise a few issues: On May 8, 2012, at 12:56 PM, Wesley Bland wrote: > We've completed the new draft of the FT chapter. I've attached just the FT chapter to this email. The rest of the document can be found on the ticket (https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323) I think this document is a bit hard to read - since the changes that are marked "ticket 0" (I assume those are all new changes) are mixed into existing changes for the main ticket, it is very hard to read what has actually been changed and what not. There is even one place where there is a ticket 0 in a ticket 0 change. Would it be possible to generate a cleaned-up ticket with just the changes? In general, though, it strikes me that some of those are non ticket 0 changes. In particular, on page 538 and 539, the new document replaces half a page of text. I think this has to be re-read, despite the obvious consequences. In addition, there are other changes that touch the semantics that make me a bit concerned. > > In this new version, we've clarified many of the points raised at the forum meetings and on the recent conference calls. Here's a list: > > - MPI_COMM_INVALIDATE is now MPI_COMM_REVOKE due to naming conflicts (along with the RMA and I/O versions) > - MPI_COMM_AGREEMENT is now MPI_COMM_AGREE to follow naming conventions > - A paragraph was added to 17.1 (Introduction) to clarify that FT is optional, but implementing the new FT functions is not. > - Section 17.2.2 (Point-to-Point and Collective Communication) has been reworked to try to make it more clean when to return MPI_ERR_PENDING and MPI_ERR_PROC_FAILED. > - The non-rooted collective discussion in the advice to users in Section 17.2.2 has been changed to make it more clear when collective operations might have non-uniform return values. what is exactly meant with "in which all processes contribute to the result and all processes receive the result" - what is "the result"? This doesn't really describe an alltoall, since any process only receives part of a "result". Perhaps this should be reformulated to "in which all processes contribute data and all processes receive data"? This probably doesn't catch MPI_Scan, though. > - The definition of MPI_COMM_REVOKE has been clarified to further define that the function is not collective and does not have a matching call on remote processes. > > If you see more changes that you think would be helpful, please let us know either on this list or on the conference call tomorrow. Martin > > Thanks, > Wesley Bland > <ftchapter.pdf>_______________________________________________ > mpi3-ft mailing list > mpi3-ft(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft ________________________________________________________________________ Martin Schulz, schulzm(a)llnl.gov, http://people.llnl.gov/schulzm CASC @ Lawrence Livermore National Laboratory, Livermore, USA

1 0

[Mpi3-ft] Updated FT Chapter
by Wesley Bland 08 May '12

08 May '12

We've completed the new draft of the FT chapter. I've attached just the FT chapter to this email. The rest of the document can be found on the ticket ( https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323) In this new version, we've clarified many of the points raised at the forum meetings and on the recent conference calls. Here's a list: - MPI_COMM_INVALIDATE is now MPI_COMM_REVOKE due to naming conflicts (along with the RMA and I/O versions) - MPI_COMM_AGREEMENT is now MPI_COMM_AGREE to follow naming conventions - A paragraph was added to 17.1 (Introduction) to clarify that FT is optional, but implementing the new FT functions is not. - Section 17.2.2 (Point-to-Point and Collective Communication) has been reworked to try to make it more clean when to return MPI_ERR_PENDING and MPI_ERR_PROC_FAILED. - The non-rooted collective discussion in the advice to users in Section 17.2.2 has been changed to make it more clear when collective operations might have non-uniform return values. - The definition of MPI_COMM_REVOKE has been clarified to further define that the function is not collective and does not have a matching call on remote processes. If you see more changes that you think would be helpful, please let us know either on this list or on the conference call tomorrow. Thanks, Wesley Bland

1 0

[Mpi3-ft] FT WG Teleconference Wednesday @ Noon EDT
by Josh Hursey 02 May '12

02 May '12

Reminder that we have a teleconf Wednesday at Noon EDT. Date: May 2, 2012 Time: Noon EDT/New York Dial-in information: 877-801-8130 Code: 1044056 Agenda: - Teleconf Notes - RMA/FT - April 25, 2012 - Tools/FT - April 30, 2012 - IO/FT - May 1, 2012 - User-Level Failure Mitigation Proposal https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323 - Timeline to vote: - 5/7: Finish ticket 0 edits to text (UTK) - 5/7: Include new fortran interfaces in text (ORNL) - 5/14: Start outlining slides for MPI Forum (George/all) - 5/14: Circulate final ticket 0 edited text to MPI Forum (all) - 5/21: Make implementation available (UTK/ORNL) - 5/21: Finish draft of slides for MPI Forum (George/all) - 5/28: MPI Forum meeting (Vote 5/30) - Implementation progress - Dynamics sub-ticket https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/327 - I/O sub-ticket https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/326 - RMA sub-ticket https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/325 - Other items? Additional Agenda Items [as time permits]: - FT Query Interface (Josh) https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/336 - FATAL scope clarification (Josh) https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/324 - FA-MPI Alternative Proposal (Tony) http://lists.mpi-forum.org/mpi3-ft/2012/02/1048.php - Piggybacking (Martin S.) https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/Piggybacking - Process Recovery (Josh H.) https://svn.mpi-forum.org/trac/mpi-forum-web/wiki/ft/process_recovery_2 - Error to Fault changes proposal (Darius B.) http://lists.mpi-forum.org/mpi3-ft/2011/06/0773.php - Network Channel Failure http://lists.mpi-forum.org/mpi3-ft/2011/06/0730.php Next Meetings: * May 9, 2012 (Teleconf) * May 16, 2012 (Teleconf - If needed) * May 23, 2012 (Teleconf) * May 28, 2012 (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] Joint IO/FT Teleconf
by Josh Hursey 01 May '12

01 May '12

I just posted the current draft of the proposal text to the ticket. Not too much has changed except the name of a few functions, per the recommendation of the forum. https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323 https://svn.mpi-forum.org/trac/mpi-forum-web/attachment/ticket/323/2012-05-… We'll use this document in our discussion today. -- Josh On Tue, May 1, 2012 at 1:30 PM, Josh Hursey <jjhursey(a)open-mpi.org> wrote: > Just a reminder that we are meeting today at 4 pm eastern. > > -- Josh > > On Thu, Apr 19, 2012 at 4:17 PM, Josh Hursey <jjhursey(a)open-mpi.org> wrote: >> It looks like May 1 from 4-5 pm Eastern works best for everyone that >> responded. Below are the teleconf details: >> ï¿½Date: May 1, 2012 >> ï¿½Time: 4 pm Eastern/New York >> ï¿½Dial-in information: 877-801-8130 >> ï¿½Code: 1044056 >> >> -- Josh >> >> On Tue, Apr 17, 2012 at 10:44 AM, Josh Hursey <jjhursey(a)open-mpi.org> wrote: >>> The FT working group would like to have a joint teleconf with the IO >>> working group to discuss the fault tolerance proposal. >>> ï¿½- https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323 >>> ï¿½- https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/326 >>> >>> A Doodle Poll link is below to schedule: >>> ï¿½- http://www.doodle.com/8iu4nryqiahfa3d9 >>> >>> If you are interested in participating in this teleconf, fill out the >>> poll by 2 pm Eastern on Thursday, April 19. >>> >>> Thanks, >>> Josh >>> >>> -- >>> Joshua Hursey >>> Postdoctoral Research Associate >>> Oak Ridge National Laboratory >>> http://users.nccs.gov/~jjhursey >> >> >> >> -- >> Joshua Hursey >> Postdoctoral Research Associate >> Oak Ridge National Laboratory >> http://users.nccs.gov/~jjhursey > > > > -- > Joshua Hursey > Postdoctoral Research Associate > Oak Ridge National Laboratory > http://users.nccs.gov/~jjhursey -- Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey

1 0

Re: [Mpi3-ft] Joint IO/FT Teleconf
by Josh Hursey 01 May '12

01 May '12

Just a reminder that we are meeting today at 4 pm eastern. -- Josh On Thu, Apr 19, 2012 at 4:17 PM, Josh Hursey <jjhursey(a)open-mpi.org> wrote: > It looks like May 1 from 4-5 pm Eastern works best for everyone that > responded. Below are the teleconf details: > ï¿½Date: May 1, 2012 > ï¿½Time: 4 pm Eastern/New York > ï¿½Dial-in information: 877-801-8130 > ï¿½Code: 1044056 > > -- Josh > > On Tue, Apr 17, 2012 at 10:44 AM, Josh Hursey <jjhursey(a)open-mpi.org> wrote: >> The FT working group would like to have a joint teleconf with the IO >> working group to discuss the fault tolerance proposal. >> ï¿½- https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/323 >> ï¿½- https://svn.mpi-forum.org/trac/mpi-forum-web/ticket/326 >> >> A Doodle Poll link is below to schedule: >> ï¿½- http://www.doodle.com/8iu4nryqiahfa3d9 >> >> If you are interested in participating in this teleconf, fill out the >> poll by 2 pm Eastern on Thursday, April 19. >> >> Thanks, >> Josh >> >> -- >> Joshua Hursey >> Postdoctoral Research Associate >> Oak Ridge National Laboratory >> http://users.nccs.gov/~jjhursey > > > > -- > Joshua Hursey > Postdoctoral Research Associate > Oak Ridge National Laboratory > http://users.nccs.gov/~jjhursey -- Joshua Hursey Postdoctoral Research Associate Oak Ridge National Laboratory http://users.nccs.gov/~jjhursey

1 0