Did you run dmesg on the node which executed the rank failed? I don't see any error message from your log file.

 

Anyway, let's try with MPICH DEBUG messages. As described in https://wiki.mpich.org/mpich/index.php/Debug_Event_Logging, first configure MPICH with "--enable-g=dbg,log" to enable event logging and built it. For example,

$ ./configure --prefix=$INSTALL_DIR --enable-g=dbg,log

$ ./make -j8

$ ./make install

 

Then, execute your application with additional options, "-mpich-dbg=file -mpich-dbg-class=all -mpich-dbg-level=verbose." With the recent MPICH, you don't need to use mpdboot. Just execute your application with mpiexec like:

mpiexec -machinefile nodes -np $NUMPROCS $EXEC -mpich-dbg=file -mpich-dbg-class=all -mpich-dbg-level=verbose

 

This will generate log files, such as dbg0-xxxx.log, dbg1-xxxx.log, etc. Since the log files may be large, it would be difficult to send them to us. If so, please take a look at each file, find error messages, and send the error message to us.

 

And, if you can reproduce the same error with small code, it will be really helpful to figure out your problem. Could you write a simplified code and send us the code?

 

Regards,

Sangmin

 

 

On Sep 15, 2014, at 10:58 AM, Abhishek Bhat <abhat@trinityconsultants.com> wrote:

Sangmin,

 

Please see attached the logs from demsg.  I apologies but I am not a computer expert so it’s all Greek to me.. Can you please see if you can find any error or reason for the failure?

 

 

Thank You

Abhishek

 

………………………………………………………………………………………………….

Abhishek Bhat, PhD, EPI,
Senior Consultant

 

 

From: Seo, Sangmin [mailto:sseo@anl.gov] 
Sent: Monday, September 15, 2014 9:21 AM
To: Abhishek Bhat
Cc: <discuss@mpich.org>
Subject: Re: [mpich-discuss] Error Running MPICH for Photochemical Modeling

 

Can you run dmesg on the node of rank 1, which is killed by signal 9, after you execute your application? You can find the reason that the process is killed at the end of dmesg output, e.g., out of memory.

 

— Sangmin

 

 

On Sep 14, 2014, at 12:37 PM, Abhishek Bhat <abhat@trinityconsultants.com> wrote:

Because the application works when less intensive runs and fails for more intensive runs, it is likely that the application is requesting too many resources.  When\where should I run ulimit –a and dmesg, after I get the error?  If that is true, is there any way to change the environment in MPI to increase the capacity so that the increased resources can be accommodated?

 

If I run it in new terminal – here is what I get

 

core file size          (blocks, -c) 0

data seg size           (kbytes, -d) unlimited

scheduling priority             (-e) 0

file size               (blocks, -f) unlimited

pending signals                 (-i) 250598

max locked memory       (kbytes, -l) 64

max memory size         (kbytes, -m) unlimited

open files                      (-n) 1024

pipe size            (512 bytes, -p) 8

POSIX message queues     (bytes, -q) 819200

real-time priority              (-r) 0

stack size              (kbytes, -s) 10240

cpu time               (seconds, -t) unlimited

max user processes              (-u) 1024

virtual memory          (kbytes, -v) unlimited

file locks                      (-x) unlimited

 

in my job, I try to set the stack size to – unlimited but I guess it is not working.

 

Let me know.  Thank you for all the help.

Abhishek

………………………………………………………………………………………………….

Abhishek Bhat, PhD, EPI,
Senior Consultant

 

 

From: Seo, Sangmin [mailto:sseo@anl.gov] 
Sent: Sunday, September 14, 2014 11:16 AM
To: <discuss@mpich.org>
Subject: Re: [mpich-discuss] Error Running MPICH for Photochemical Modeling

 

Abhishek,

 

Signal 9 is caused by many reasons, e.g., CPU time, out of memory, etc., but it is mostly because the application requests too many resources. You can check the environment settings with ulimit -a. And, you may find some information about your error from dmesg.

 

Thanks,

Sangmin

 

 

On Sep 12, 2014, at 5:51 PM, Abhishek Bhat <abhat@trinityconsultants.com> wrote:

Sangmin.

 

I updated to mpich3 and getting the following error

 

Fatal error in MPI_Recv: A process has failed, error stack:

MPI_Recv(187).............: MPI_Recv(buf=0x7fff93840c30, count=644490, MPI_REAL, src=1, tag=14131, MPI_COMM_WORLD, status=0x7fff94444f20) failed

dequeue_and_set_error(865): Communication error with rank 1

rank 1 in job 1  dfw-camx_55000   caused collective abort of all ranks

  exit status of rank 1: killed by signal 9

 

Same situation, successful runs for smaller resource runs and for up to 7 processes.  Error at more than 7.  Here is the mpich command I am using to run from my job file…

 

cat << ieof > nodes

dfw-camx:1

dfw-camx-n1:1

dfw-camx-n2:1

dfw-camx-n3:1

dfw-camx-n4:1

dfw-camx-n5:1

dfw-camx-n6:1

dfw-camx-n7:1

ieof

set NUMPROCS = 8

set RING = `wc -l nodes | awk '{print $1}'`

mpdboot -n $RING -f nodes –verbose

 

if( ! { mpiexec -machinefile nodes -np $NUMPROCS $EXEC } ) then

   mpdallexit

   exit

endif

 

 

For a successful run the NUMPROCS has to be < = 7.

 

Any help is much appreciated.

 

Thank You

Abhishek

………………………………………………………………………………………………….

Abhishek Bhat, PhD, EPI,
Senior Consultant

 

 

From: Seo, Sangmin [mailto:sseo@anl.gov] 
Sent: Friday, September 12, 2014 1:11 PM
To: <discuss@mpich.org>
Subject: Re: [mpich-discuss] Error Running MPICH for Photochemical Modeling

 

Hi Abhishek,

 

Can you try with the recent MPICH release to see if the same error happens? You can download the recent release, 3.1.2, from http://www.mpich.org/downloads/.

 

Thanks,

Sangmin

 

 

On Sep 12, 2014, at 12:59 PM, Abhishek Bhat <abhat@trinityconsultants.com> wrote:

I am running a photochemical modeling on Linux cluster (CentOS_64 bit) with 1 master and 8 slave nodes with quad core (intel i7) on each node.  I have two scenarios, in first scenario, I am running less data intensive run on all 8 nodes (NUMPROCS = 9) and the run will go fine.  When running same configuration for a more intense run, I am getting following error.

 

Fatal error in MPI_Recv: Other MPI error, error stack:

MPI_Recv(187).....................: MPI_Recv(buf=0x7fff989d53b0, count=644490, MPI_REAL, src=1, tag=14131, MPI_COMM_WORLD, status=0x7fff995d96a0) failed

MPIDI_CH3I_Progress(150)..........:

MPID_nem_mpich2_blocking_recv(948):

MPID_nem_tcp_connpoll(1720).......:

state_commrdy_handler(1556).......:

MPID_nem_tcp_recv_handler(1446)...: socket closed

rank 1 in job 1  dfw-camx_55000   caused collective abort of all ranks

  exit status of rank 1: killed by signal 9

 

If I run the program with smaller nodes (smaller than 7 NUMPROCS) the run goes fine.

 

It appears that the rank 1 (my first node) is collectively causing all the ranks, but I could identify why.  I tried following solutions –

 

1.       Increased master memory to 32 gb

2.       Increased all nodes memory to 32 gb

3.       Exchanged the rank 1 to different node in the parallel.

 

In all situations, I am getting this error.  Surprisingly, when I am running smaller (less data intensive runs), I am not getting this error even if I increase the NUMPROCS to 32 processes.

 

Any help will be highly appreciated.

 

I am running mpich 1.4

 

Thank You
Abhishek

………………………………………………………………………………………………….

Abhishek Bhat, PhD, EPI,
Senior Consultant

 

Trinity Consultants

12770 Merit Drive, Suite 900  |  Dallas, Texas 75251

Office:  972-661-8100|  Mobile:  806-281-7617

Email:  abhat@trinityconsultants.com  |  LinkedIn: www.linkedin.com/in/abhattrinityconsultants

 

Stay current on environmental issues.  Subscribe today to receive Trinity's free Environmental Quarterly.

Learn about Trinity’s courses for environmental professionals.

 

<image001.gif>    <image002.gif>    <image003.gif>    <image004.gif>

 

<image005.jpg>

 

_________________________________________________________________________

The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material. Any review, retransmission, dissemination or other use of, or
taking of any action in reliance upon, this information by persons or
entities other than the intended recipient is prohibited. If you received
this in error, please contact the sender and delete the material from any
computer.
_________________________________________________________________________
_______________________________________________
discuss mailing list     discuss@mpich.org
To manage subscription options or unsubscribe:
https://lists.mpich.org/mailman/listinfo/discuss

 

_________________________________________________________________________

The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material. Any review, retransmission, dissemination or other use of, or
taking of any action in reliance upon, this information by persons or
entities other than the intended recipient is prohibited. If you received
this in error, please contact the sender and delete the material from any
computer.
_________________________________________________________________________
_______________________________________________
discuss mailing list     discuss@mpich.org
To manage subscription options or unsubscribe:
https://lists.mpich.org/mailman/listinfo/discuss

 

_________________________________________________________________________

The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material. Any review, retransmission, dissemination or other use of, or
taking of any action in reliance upon, this information by persons or
entities other than the intended recipient is prohibited. If you received
this in error, please contact the sender and delete the material from any
computer.
_________________________________________________________________________

 

_________________________________________________________________________

The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material. Any review, retransmission, dissemination or other use of, or
taking of any action in reliance upon, this information by persons or
entities other than the intended recipient is prohibited. If you received
this in error, please contact the sender and delete the material from any
computer.
_________________________________________________________________________
<dmesg3.txt><messages.txt>

 

_______________________________________________
discuss mailing list     discuss@mpich.org
To manage subscription options or unsubscribe:
https://lists.mpich.org/mailman/listinfo/discuss