September 2008 - mpiwg-ft - mailman3.cels.anl.gov

Re: [Mpi3-ft] Weekly con-calls
by Hideyuki Jitsumoto 24 Sep '08

24 Sep '08

Everyone, I'm sorry but something urgent came up, so I'm afraid I will not be able to attend today's telecon. I'll be able to attend next week. I'm terribly sorry, Hideyuki Jitsumoto 2008/9/24 Richard Graham <rlgraham(a)ornl.gov>: > Just a reminder about tomorrow's call. > Rich > > > On 9/16/08 5:05 PM, "Richard Graham" <rlgraham(a)ornl.gov> wrote: > > Lets resume these meetings on Wed next week, 9/24/2008, noon-2pm Eastern > time ï¿½ moving this an hours later does not work for the people from Japan. > What we want to talk about next time are specific user use-cases. We have > talked so far in generalities, but now need to get to specifics in term of > how apps actually want to use this functionality. As an example, take the > case of client/server apps, with a failure in on of the "clients". If the > client is member of the remote group in the intercommunicator, fully > defining error scenarios in the current MPI-2 dynamics should be sufficient. > However, if a member of the failed client is in the some comm_world as the > sever then we need to define new semantics. These are the sort of details > we need to get to, so please bring as much input as you can from the user > communities you work with. > > Call details: > Wed - 12-2 pm EDT, 5-7 pm GMT > US Toll Free number: 877-801-8130 > Toll number: 1-203-692-8690 > Access Code: 1044056 > > Rich > > ________________________________ > _______________________________________________ > mpi3-ft mailing list > mpi3-ft(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft > > _______________________________________________ > mpi3-ft mailing list > mpi3-ft(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft > > -- Sincerely Yours, Hideyuki Jitsumoto (jitumoto(a)matsulab.is.titech.ac.jp) Tokyo Institute of Technology Global Scientific Information and Computing center (Matsuoka Lab.)

1 0

Re: [Mpi3-ft] Weekly con-calls
by Greg Bronevetsky 23 Sep '08

23 Sep '08

I won't be able to make it to this week's telecon but future telecons will be fine. Greg Bronevetsky Post-Doctoral Researcher 1028 Building 451 Lawrence Livermore National Lab (925) 424-5756 bronevetsky1(a)llnl.gov At 02:22 PM 9/23/2008, Richard Graham wrote: >Just a reminder about tomorrows call. >Rich > > >On 9/16/08 5:05 PM, "Richard Graham" ><<rlgraham(a)ornl.htm>rlgraham(a)ornl.gov> wrote: > >Lets resume these meetings on Wed next week, >9/24/2008, noon-2pm Eastern time moving this >an hours later does not work for the people from >Japan. What we want to talk about next time are >specific user use-cases. We have talked so far >in generalities, but now need to get to >specifics in term of how apps actually want to >use this functionality. As an example, take the >case of client/server apps, with a failure in on >of the clients. If the client is member of >the remote group in the intercommunicator, fully >defining error scenarios in the current MPI-2 >dynamics should be sufficient. However, if a >member of the failed client is in the some >comm_world as the sever then we need to define >new semantics. These are the sort of details >we need to get to, so please bring as much input >as you can from the user communities you work with. > >Call details: >Wed - 12-2 pm EDT, 5-7 pm GMT > US Toll Free number: 877-801-8130 > Toll number: 1-203-692-8690 > Access Code: 1044056 > >Rich > > >_______________________________________________ >mpi3-ft mailing list ><mpi3-ft(a)lists.mpi-forum.htm>mpi3-ft(a)lists.mpi-forum.org >http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft > >_______________________________________________ >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] Weekly con-calls
by Richard Graham 23 Sep '08

23 Sep '08

Just a reminder about tomorrow¹s call. Rich On 9/16/08 5:05 PM, "Richard Graham" <rlgraham(a)ornl.gov> wrote: > Lets resume these meetings on Wed next week, 9/24/2008, noon-2pm Eastern time > moving this an hours later does not work for the people from Japan. What we > want to talk about next time are specific user use-cases. We have talked so > far in generalities, but now need to get to specifics in term of how apps > actually want to use this functionality. As an example, take the case of > client/server apps, with a failure in on of the ³clients². If the client is > member of the remote group in the intercommunicator, fully defining error > scenarios in the current MPI-2 dynamics should be sufficient. However, if a > member of the failed client is in the some comm_world as the sever then we > need to define new semantics. These are the sort of details we need to get > to, so please bring as much input as you can from the user communities you > work with. > > Call details: > Wed - 12-2 pm EDT, 5-7 pm GMT > US Toll Free number: 877-801-8130 > Toll number: 1-203-692-8690 > Access Code: 1044056 > > Rich > > > _______________________________________________ > 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] Weekly con-calls
by Greg Bronevetsky 22 Sep '08

22 Sep '08

>Lets resume these meetings on Wed next week, 9/24/2008, noon-2pm Eastern >time ï¿½ moving this an hours later does not work for the people from Japan. >What we want to talk about next time are specific user use-cases. We have >talked so far in generalities, but now need to get to specifics in term of >how apps actually want to use this functionality. As an example, take the >case of client/server apps, with a failure in on of the ï¿½clientsï¿½. If the >client is member of the remote group in the intercommunicator, fully >defining error scenarios in the current MPI-2 dynamics should be sufficient Lets start talking about this scenario. The first thing to talk about is the functionality that may be desired by the application. There are two cases to consider. When the client is not expecting a response from the server, its failure is irrelevant to the server and the server doesn't want to be informed of the failure. However, if the client fails with pending communication from the server, the server does want to be informed of the failure, if only to free the internal state associated with this client. As such, the application may desire one of two things. First, it may wish MPI to inform it every time it tries to communicate with a failed process. This way the server is notified during its next send or receive operation to/from the client and is thus able to perform the appropriate cleanup operations. Furthermore, if the server wants to be able to respond as quickly as possible, it can ask MPI to tell it via a callback when the client fails. We've already discussed both types of interfaces in the group. More specifically, the MPI should inform the application either during the application's next call that uses the above intercommunicator or invoke the callback associated with this intercommunicator. Looking at the problem at a lower level, we can consider the various low-level problems that may cause client "failure" and what MPI must do to support the application's recovery efforts. The first possibility is that the client node fails. MPI can detect this event by using heart-beat messages or by noticing that a message sent to the client has timed out. In this case MPI should use one of the above APIs to inform the other processes that further communication to/from the client will fail. Another possibility is that some network link fails, disabling communication between this client and this server but allowing communication between other rank pairs. In this case MPI may inform the application about the pairs of ranks that can no longer communicate. Another option would be for MPI to kill half the processes that are affected by this problem, ensuring that all the remaining processes can communicate freely with each other. It would then inform the application of these process deaths as above. The final possibility is a network partition where a subset of processes cannot communicate with the remaining processes. MPI would treat this case as one of the subsets failing and kill all the processes in this subset. It would then inform the processes in the other subset of these deaths. The above is essentially a summary of what we've talked about thus far, as applied to this example. This is good, since this suggests that we understand the problem well enough, at least for this purpose. A good question then is, can anybody think of any additional details that we need to think about? Greg Bronevetsky Post-Doctoral Researcher 1028 Building 451 Lawrence Livermore National Lab (925) 424-5756 bronevetsky1(a)llnl.gov

1 0

[Mpi3-ft] Weekly con-calls
by Richard Graham 16 Sep '08

16 Sep '08

Lets resume these meetings on Wed next week, 9/24/2008, noon-2pm Eastern time moving this an hours later does not work for the people from Japan. What we want to talk about next time are specific user use-cases. We have talked so far in generalities, but now need to get to specifics in term of how apps actually want to use this functionality. As an example, take the case of client/server apps, with a failure in on of the ³clients². If the client is member of the remote group in the intercommunicator, fully defining error scenarios in the current MPI-2 dynamics should be sufficient. However, if a member of the failed client is in the some comm_world as the sever then we need to define new semantics. These are the sort of details we need to get to, so please bring as much input as you can from the user communities you work with. Call details: Wed - 12-2 pm EDT, 5-7 pm GMT US Toll Free number: 877-801-8130 Toll number: 1-203-692-8690 Access Code: 1044056 Rich

1 0

Re: [Mpi3-ft] Weekly meetings
by Solt, David George 16 Sep '08

16 Sep '08

What is the status of the new meeting time? I would like to start attending, but I do not see any information on the new time. Thanks, Dave Solt -----Original Message----- From: mpi3-ft-bounces(a)lists.mpi-forum.org [mailto:[email protected]] On Behalf Of Josh Hursey Sent: Friday, September 05, 2008 8:57 PM To: MPI 3.0 Fault Tolerance and Dynamic Process Control working Group Subject: Re: [Mpi3-ft] Weekly meetings I have had to miss the past few meetings due to various conflicts (I have been following the discussion on the list though). Unfortunately it looks like the Friday time will no longer work for me, so I am in support of moving the day of the teleconf. However, if we move the meeting to Wednesday I will not be able to attend at the standard time since I have a conflict from 11:30 - 1:00 EST. If we shifted it to be from 1 to 3 pm EST on Wednesday that would work much better for me. Just a suggestion though. Cheers, Josh On Sep 5, 2008, at 5:12 PM, Richard Graham wrote: > At the meeting here in Dublin this week we decided to move the weekly > meeting from Friday to Wed - same time. Please let me know if this > causes you a problem. We will NOT have a call this coming week, the > week of 9/8, as several of us are at Euro PVM./MPI, but will resume > the following week. > > Rich > _______________________________________________ > mpi3-ft mailing list > mpi3-ft(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft _______________________________________________ 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] Weekly meetings
by Josh Hursey 05 Sep '08

05 Sep '08

I have had to miss the past few meetings due to various conflicts (I have been following the discussion on the list though). Unfortunately it looks like the Friday time will no longer work for me, so I am in support of moving the day of the teleconf. However, if we move the meeting to Wednesday I will not be able to attend at the standard time since I have a conflict from 11:30 - 1:00 EST. If we shifted it to be from 1 to 3 pm EST on Wednesday that would work much better for me. Just a suggestion though. Cheers, Josh On Sep 5, 2008, at 5:12 PM, Richard Graham wrote: > At the meeting here in Dublin this week we decided to move the > weekly meeting from Friday to Wed ï¿½ same time. Please let me know > if this causes you a problem. We will NOT have a call this coming > week, the week of 9/8, as several of us are at Euro PVM./MPI, but > will resume the following week. > > Rich > _______________________________________________ > mpi3-ft mailing list > mpi3-ft(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpi3-ft

1 0

[Mpi3-ft] Weekly meetings
by Richard Graham 05 Sep '08

05 Sep '08

At the meeting here in Dublin this week we decided to move the weekly meeting from Friday to Wed same time. Please let me know if this causes you a problem. We will NOT have a call this coming week, the week of 9/8, as several of us are at Euro PVM./MPI, but will resume the following week. Rich

1 0

[Mpi3-ft] Error reporting
by Richard Graham 04 Sep '08

04 Sep '08

Here is the current document dealing with error reporting in MPI. Rich

1 0