- devel - mailman3.cels.anl.gov

old old MPICH history now on line
by Rob Latham 15 Jul '15

15 Jul '15

Did you ever run 'git blame' to find out why some part of MPICH does what it does, only to find the commit responsible was 6a1cbdcfc42 "Added all components in the right places."? I'm personally surprised how often I care about changes made more than 8 years ago. If you, like me, sometimes find yourself needing more history, you can now -- ta-da -- consult http://git.mpich.org/mpich-CVS.git/ When we converted from CVS to subversion back in 2007, the conversion tools could not handle some oddities in our CVS repository, necessitating a wipeoput of history. I guess in the intervening 8 years the conversion tools are more tolerant (once you repack your repository to remove any modules-containing-modules). This is still a disconnected tree from today's MPICH development, but now you can browse the old history with the tools we are familiar with today (no need to build/install cvsps or remember old CVS commands). ==rob -- Rob Latham Mathematics and Computer Science Division Argonne National Lab, IL USA

1 0

Re: [mpich-devel] about MPI hydra process management
by unix haha 01 Jul '15

01 Jul '15

^__^ Many thanks! Pavan On Wed, Jul 1, 2015 at 7:59 PM, Balaji, Pavan <balaji(a)anl.gov> wrote: > > We are not very good at names. :-) > > "persist" is for a "persistent daemon", which makes no sense, since sshd > is a persistent too. > > "bss" stands for "bootstrap server", and the name conflict is > unfortunate. > > -- Pavan > > From: unix haha > Reply-To: "devel(a)mpich.org" > Date: Tuesday, June 30, 2015 at 9:35 PM > To: "devel(a)mpich.org" > Subject: Re: [mpich-devel] about MPI hydra process management > > Thanks, Balaji. > > I took a quick look on persist bootstrap. It seems what I want and I will > try it. > > one thing puzzles me is what's "persist" means here for hydra. what's > means for hydra_bss_persist in Makefile.mk under hydra/tools/bootstrap/? > the bss is the .bss section in executable file? > > Thanks! > > On Wed, Jul 1, 2015 at 7:29 AM, Balaji, Pavan <balaji(a)anl.gov> wrote: > >> >> Hydra supports a bunch of daemon processes, including ssh. There is a >> daemon process included with MPICH called "persist", which you could use >> for this purpose, but it's not really maintained. You could try it. >> >> -- Pavan >> >> From: unix haha >> Reply-To: "devel(a)mpich.org" >> Date: Tuesday, June 30, 2015 at 6:18 PM >> To: "devel(a)mpich.org" >> Subject: [mpich-devel] about MPI hydra process management >> >> Hi, all: >> >> I'm working on port the MPICH2 3.1.4 to an embedded environment. I >> met some problems for hydra process management, Could anyone help me? >> >> 1. For the old MPD, before the mpirun, there's a daemon process >> running on all computer nodes. But I don't see what's hydra's daemon >> process in the current MPI document. Could you give me some hints? >> >> 2. As we run the MPICH2 in a resource constraint environment, the >> SSH or the other linux-based launcher may not be used. Does it exist any >> simple launcher running on the computer nodes before the mpiexec? >> >> 3. What's my understanding for hydra is: once the Launcher process >> running on all computer nodes, I can use the mpiexec on the control node to >> start the parallel jobs? If so, the problem is: Could I start the Launcher >> process manually if there's no linux-based launcher? >> >> Thanks! >> >> _______________________________________________ >> To manage subscription options or unsubscribe: >> https://lists.mpich.org/mailman/listinfo/devel >> > > > _______________________________________________ > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/devel >

1 0

Re: [mpich-devel] about MPI hydra process management
by Balaji, Pavan 01 Jul '15

01 Jul '15

We are not very good at names. :-) "persist" is for a "persistent daemon", which makes no sense, since sshd is a persistent too. "bss" stands for "bootstrap server", and the name conflict is unfortunate. -- Pavan From: unix haha Reply-To: "devel(a)mpich.org<mailto:[email protected]>" Date: Tuesday, June 30, 2015 at 9:35 PM To: "devel(a)mpich.org<mailto:[email protected]>" Subject: Re: [mpich-devel] about MPI hydra process management Thanks, Balaji. I took a quick look on persist bootstrap. It seems what I want and I will try it. one thing puzzles me is what's "persist" means here for hydra. what's means for hydra_bss_persist in Makefile.mk under hydra/tools/bootstrap/? the bss is the .bss section in executable file? Thanks! On Wed, Jul 1, 2015 at 7:29 AM, Balaji, Pavan <balaji(a)anl.gov<mailto:[email protected]>> wrote: Hydra supports a bunch of daemon processes, including ssh. There is a daemon process included with MPICH called "persist", which you could use for this purpose, but it's not really maintained. You could try it. -- Pavan From: unix haha Reply-To: "devel(a)mpich.org<mailto:[email protected]>" Date: Tuesday, June 30, 2015 at 6:18 PM To: "devel(a)mpich.org<mailto:[email protected]>" Subject: [mpich-devel] about MPI hydra process management Hi, all: I'm working on port the MPICH2 3.1.4 to an embedded environment. I met some problems for hydra process management, Could anyone help me? 1. For the old MPD, before the mpirun, there's a daemon process running on all computer nodes. But I don't see what's hydra's daemon process in the current MPI document. Could you give me some hints? 2. As we run the MPICH2 in a resource constraint environment, the SSH or the other linux-based launcher may not be used. Does it exist any simple launcher running on the computer nodes before the mpiexec? 3. What's my understanding for hydra is: once the Launcher process running on all computer nodes, I can use the mpiexec on the control node to start the parallel jobs? If so, the problem is: Could I start the Launcher process manually if there's no linux-based launcher? Thanks! _______________________________________________ To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/devel

1 0

Re: [mpich-devel] about MPI hydra process management
by unix haha 30 Jun '15

30 Jun '15

Thanks, Balaji. I took a quick look on persist bootstrap. It seems what I want and I will try it. one thing puzzles me is what's "persist" means here for hydra. what's means for hydra_bss_persist in Makefile.mk under hydra/tools/bootstrap/? the bss is the .bss section in executable file? Thanks! On Wed, Jul 1, 2015 at 7:29 AM, Balaji, Pavan <balaji(a)anl.gov> wrote: > > Hydra supports a bunch of daemon processes, including ssh. There is a > daemon process included with MPICH called "persist", which you could use > for this purpose, but it's not really maintained. You could try it. > > -- Pavan > > From: unix haha > Reply-To: "devel(a)mpich.org" > Date: Tuesday, June 30, 2015 at 6:18 PM > To: "devel(a)mpich.org" > Subject: [mpich-devel] about MPI hydra process management > > Hi, all: > > I'm working on port the MPICH2 3.1.4 to an embedded environment. I > met some problems for hydra process management, Could anyone help me? > > 1. For the old MPD, before the mpirun, there's a daemon process > running on all computer nodes. But I don't see what's hydra's daemon > process in the current MPI document. Could you give me some hints? > > 2. As we run the MPICH2 in a resource constraint environment, the > SSH or the other linux-based launcher may not be used. Does it exist any > simple launcher running on the computer nodes before the mpiexec? > > 3. What's my understanding for hydra is: once the Launcher process > running on all computer nodes, I can use the mpiexec on the control node to > start the parallel jobs? If so, the problem is: Could I start the Launcher > process manually if there's no linux-based launcher? > > Thanks! > > _______________________________________________ > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/devel >

1 0

Re: [mpich-devel] about MPI hydra process management
by Balaji, Pavan 30 Jun '15

30 Jun '15

Hydra supports a bunch of daemon processes, including ssh. There is a daemon process included with MPICH called "persist", which you could use for this purpose, but it's not really maintained. You could try it. -- Pavan From: unix haha Reply-To: "devel(a)mpich.org<mailto:[email protected]>" Date: Tuesday, June 30, 2015 at 6:18 PM To: "devel(a)mpich.org<mailto:[email protected]>" Subject: [mpich-devel] about MPI hydra process management Hi, all: I'm working on port the MPICH2 3.1.4 to an embedded environment. I met some problems for hydra process management, Could anyone help me? 1. For the old MPD, before the mpirun, there's a daemon process running on all computer nodes. But I don't see what's hydra's daemon process in the current MPI document. Could you give me some hints? 2. As we run the MPICH2 in a resource constraint environment, the SSH or the other linux-based launcher may not be used. Does it exist any simple launcher running on the computer nodes before the mpiexec? 3. What's my understanding for hydra is: once the Launcher process running on all computer nodes, I can use the mpiexec on the control node to start the parallel jobs? If so, the problem is: Could I start the Launcher process manually if there's no linux-based launcher? Thanks!

1 0

about MPI hydra process management
by unix haha 30 Jun '15

30 Jun '15

Hi, all: I'm working on port the MPICH2 3.1.4 to an embedded environment. I met some problems for hydra process management, Could anyone help me? 1. For the old MPD, before the mpirun, there's a daemon process running on all computer nodes. But I don't see what's hydra's daemon process in the current MPI document. Could you give me some hints? 2. As we run the MPICH2 in a resource constraint environment, the SSH or the other linux-based launcher may not be used. Does it exist any simple launcher running on the computer nodes before the mpiexec? 3. What's my understanding for hydra is: once the Launcher process running on all computer nodes, I can use the mpiexec on the control node to start the parallel jobs? If so, the problem is: Could I start the Launcher process manually if there's no linux-based launcher? Thanks!

1 0

Re: [mpich-devel] MPI_Init triggers Windows Firewall
by Anh Vo 30 Jun '15

30 Jun '15

Hi Colin, If this is MS-MPI you can set the environment variable MSMPI_DISABLE_SOCK to 1 before running the job, that should prevent the Windows Firewall from popping up Anh On Thu, Jun 25, 2015 at 12:41 PM, Antonio J. Peña <apenya(a)mcs.anl.gov> wrote: > > Sorry, MPICH on Windows is unsupported (and so it is the use of MPI > without MPI_Init). > > Please refer to our FAQ: > > https://wiki.mpich.org/mpich/index.php/Frequently_Asked_Questions#Q:_Why_ca… > > > On 06/25/2015 02:26 PM, Colin Potter wrote: > > Is there any way to get the MPI_Init call to not trigger the Windows > Firewall dialog? It asks for TCP and UDP access. This is in an application > being run standalone, without mpiexec. > > If I comment out MPI_Init, it does not trigger the dialog. > > > _______________________________________________ > To manage subscription options or unsubscribe:https://lists.mpich.org/mailman/listinfo/devel > > > > -- > Antonio J. Peña > Postdoctoral Appointee > Mathematics and Computer Science Division > Argonne National Laboratory > 9700 South Cass Avenue, Bldg. 240, Of. 3148 > Argonne, IL 60439-4847apenya(a)mcs.anl.govwww.mcs.anl.gov/~apenya > > > _______________________________________________ > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/devel >

1 0

Re: [mpich-devel] MPI_Init triggers Windows Firewall
by Antonio J. Peña 25 Jun '15

25 Jun '15

Sorry, MPICH on Windows is unsupported (and so it is the use of MPI without MPI_Init). Please refer to our FAQ: https://wiki.mpich.org/mpich/index.php/Frequently_Asked_Questions#Q:_Why_ca… On 06/25/2015 02:26 PM, Colin Potter wrote: > Is there any way to get the MPI_Init call to not trigger the Windows > Firewall dialog? It asks for TCP and UDP access. This is in an > application being run standalone, without mpiexec. > > If I comment out MPI_Init, it does not trigger the dialog. > > > _______________________________________________ > To manage subscription options or unsubscribe: > https://lists.mpich.org/mailman/listinfo/devel -- Antonio J. Peña Postdoctoral Appointee Mathematics and Computer Science Division Argonne National Laboratory 9700 South Cass Avenue, Bldg. 240, Of. 3148 Argonne, IL 60439-4847 apenya(a)mcs.anl.gov www.mcs.anl.gov/~apenya

1 0

MPI_Init triggers Windows Firewall
by Colin Potter 25 Jun '15

25 Jun '15

Is there any way to get the MPI_Init call to not trigger the Windows Firewall dialog? It asks for TCP and UDP access. This is in an application being run standalone, without mpiexec. If I comment out MPI_Init, it does not trigger the dialog.

1 0

Re: [mpich-devel] MPI_Win leaking memory
by Zhao, Xin 08 Jun '15

08 Jun '15

Hi Lisandro, You can track the progress of this bug here: https://trac.mpich.org/projects/mpich/ticket/2273 Thanks, Xin ________________________________________ From: Zhao, Xin [xinzhao3(a)illinois.edu] Sent: Friday, June 05, 2015 10:14 AM To: devel(a)mpich.org Subject: Re: [mpich-devel] MPI_Win leaking memory Hi Lisandro, Thanks. I will look into this. Xin ________________________________________ From: Lisandro Dalcin [dalcinl(a)gmail.com] Sent: Friday, June 05, 2015 10:02 AM To: mpich-devel Subject: Re: [mpich-devel] MPI_Win leaking memory On 5 June 2015 at 15:01, Balaji, Pavan <balaji(a)anl.gov> wrote: > > Huh. I wonder why our testing is not catching this. Can you send us the exact test you used? > This is the minimal Python code able to reproduce the leak, translation to C is obvious. I would not be surprised if none of your tests have such a code path. from mpi4py import MPI win = MPI.Win.Create(MPI.BOTTOM, 1, MPI.INFO_NULL, MPI.COMM_SELF) info = win.Get_info() win.Set_info(info) info.Free() win.Free() -- Lisandro Dalcin ============ Research Scientist Computer, Electrical and Mathematical Sciences & Engineering (CEMSE) Numerical Porous Media Center (NumPor) King Abdullah University of Science and Technology (KAUST) http://numpor.kaust.edu.sa/ 4700 King Abdullah University of Science and Technology al-Khawarizmi Bldg (Bldg 1), Office # 4332 Thuwal 23955-6900, Kingdom of Saudi Arabia http://www.kaust.edu.sa Office Phone: +966 12 808-0459 _______________________________________________ To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/devel _______________________________________________ To manage subscription options or unsubscribe: https://lists.mpich.org/mailman/listinfo/devel

1 0