<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style>
</head>
<body class='hmmessage'><div dir='ltr'>
I disabled the "ufw" and the error continue :-(<br><br>Miguel Angel<br><br><div>> Date: Sat, 22 Oct 2011 16:47:57 -0500<br>> From: balaji@mcs.anl.gov<br>> To: mafga74@hotmail.com<br>> CC: mpich-discuss@mcs.anl.gov<br>> Subject: Re: [mpich-discuss] MPID_nem_tcp_connpoll(1826): Communication error with rank 0: Connection refused<br>> <br>> <br>> [Please keep mpich-discuss cc'ed. I tend to ignore MPI question emails <br>> sent directly to me.]<br>> <br>> Ubuntu does install a firewall automatically, AFAIK. You can check:<br>> <br>> % sudo ufw status<br>> <br>> -- Pavan<br>> <br>> On 10/22/2011 04:38 PM, Miguel Angel Fernández wrote:<br>> > Unles Ubuntu installs a firewall automaticly, I don't understand what<br>> > can be the problem.<br>> ><br>> > Anyway, I'm going to install the WireShark (Ethereal) on Ubuntu. Maybe<br>> > this clarify the problem.<br>> ><br>> > Thanks<br>> > Miguel Angel<br>> ><br>> > > Date: Sat, 22 Oct 2011 16:20:43 -0500<br>> > > From: balaji@mcs.anl.gov<br>> > > To: mpich-discuss@mcs.anl.gov<br>> > > CC: mafga74@hotmail.com<br>> > > Subject: Re: [mpich-discuss] MPID_nem_tcp_connpoll(1826):<br>> > Communication error with rank 0: Connection refused<br>> > ><br>> > ><br>> > > It picks whichever port it can find. It's not fixed.<br>> > ><br>> > > This does look like a firewall problem. Most firewalls, keep the ssh<br>> > > port (22) open, so being able to do an ssh doesn't mean there's no<br>> > firewall.<br>> > ><br>> > > On 10/22/2011 03:46 PM, Miguel Angel Fernández wrote:<br>> > > > Hi Rajeev<br>> > > ><br>> > > > Thanks for your response but there is no any firewall among them.<br>> > > ><br>> > > > mpi0 is Ubuntu<br>> > > > mpi1 to mpi4 are Debian (minimun distribution without GUI)<br>> > > ><br>> > > > I am doing the conection among them with ssh and it is well configured<br>> > > > and properly working, I tested it.<br>> > > > I am thinking,...are you using a socket conettion to comunicate the<br>> > > > diferent processes? This could be the problem in the Ubuntu one (mpi0).<br>> > > > If this is the case, can you tell me the exactly socket you are using?<br>> > > ><br>> > > > Thanks<br>> > > > Miguel Angel<br>> > > ><br>> > > ><br>> > > > > From: thakur@mcs.anl.gov<br>> > > > > Date: Sat, 22 Oct 2011 15:30:26 -0500<br>> > > > > To: mpich-discuss@mcs.anl.gov<br>> > > > > Subject: Re: [mpich-discuss] MPID_nem_tcp_connpoll(1826):<br>> > > > Communication error with rank 0: Connection refused<br>> > > > ><br>> > > > > Make sure the 5 machines can communicate with each other, i.e., there<br>> > > > is no firewall preventing connections.<br>> > > > ><br>> > > > > Rajeev<br>> > > > ><br>> > > > > On Oct 22, 2011, at 12:36 PM, Miguel Angel Fernández wrote:<br>> > > > ><br>> > > > > > Hello everybody<br>> > > > > ><br>> > > > > > I'm trying to fix a problem that appear when I execute one of the<br>> > > > mpich2 program examples.<br>> > > > > > As you can see, if I execute a normal command there are no<br>> > > > problems. The cluster work properly.<br>> > > > > ><br>> > > > > > mpi@mpi0:~$ mpiexec -f ./mpich2-install/machinefile -n 5 hostname<br>> > > > > > mpi0<br>> > > > > > mpi2<br>> > > > > > mpi3<br>> > > > > > mpi1<br>> > > > > > mpi4<br>> > > > > > mpi@mpi0:~$<br>> > > > > ><br>> > > > > > but when I try to execute the program, the results are something<br>> > > > like this<br>> > > > > ><br>> > > > > > mpi@mpi0:~$ mpiexec -f ./mpich2-install/machinefile -n 5<br>> > > > /home/mpi/mpich2-install/workspace/Prueba/Debug/Prueba<br>> > > > > > Hello MPI World the original.<br>> > > > > > Hello MPI World the original.<br>> > > > > > Hello MPI World the original.<br>> > > > > > Hello MPI World the original.<br>> > > > > > Hello MPI World the original.<br>> > > > > > From process 0: Num processes: 5<br>> > > > > > Fatal error in MPI_Send: Other MPI error, error stack:<br>> > > > > > MPI_Send(173)..............: MPI_Send(buf=0xbfcbe268, count=26,<br>> > > > MPI_CHAR, dest=0, tag=0, MPI_COMM_WORLD) failed<br>> > > > > > MPID_nem_tcp_connpoll(1826): Communication error with rank 0:<br>> > > > Connection refused<br>> > > > > > Fatal error in MPI_Send: Other MPI error, error stack:<br>> > > > > > MPI_Send(173)..............: MPI_Send(buf=0xbfb32ca8, count=26,<br>> > > > MPI_CHAR, dest=0, tag=0, MPI_COMM_WORLD) failed<br>> > > > > > MPID_nem_tcp_connpoll(1826): Communication error with rank 0:<br>> > > > Connection refused<br>> > > > > > Fatal error in MPI_Send: Other MPI error, error stack:<br>> > > > > > MPI_Send(173)..............: MPI_Send(buf=0xbfa49e98, count=26,<br>> > > > MPI_CHAR, dest=0, tag=0, MPI_COMM_WORLD) failed<br>> > > > > > MPID_nem_tcp_connpoll(1826): Communication error with rank 0:<br>> > > > Connection refused<br>> > > > > > Fatal error in MPI_Send: Other MPI error, error stack:<br>> > > > > > MPI_Send(173)..............: MPI_Send(buf=0xbfa57538, count=26,<br>> > > > MPI_CHAR, dest=0, tag=0, MPI_COMM_WORLD) failed<br>> > > > > > MPID_nem_tcp_connpoll(1826): Communication error with rank 0:<br>> > > > Connection refused<br>> > > > > ><br>> > > > > > Do you have any idea what can be the problem?<br>> > > > > ><br>> > > > > > Thank you in advance<br>> > > > > > Miguel Angel<br>> > > > > ><br>> > > > > > _______________________________________________<br>> > > > > > mpich-discuss mailing list mpich-discuss@mcs.anl.gov<br>> > > > > > To manage subscription options or unsubscribe:<br>> > > > > > https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss<br>> > > > ><br>> > > > > _______________________________________________<br>> > > > > mpich-discuss mailing list mpich-discuss@mcs.anl.gov<br>> > > > > To manage subscription options or unsubscribe:<br>> > > > > https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss<br>> > > ><br>> > > ><br>> > > > _______________________________________________<br>> > > > mpich-discuss mailing list mpich-discuss@mcs.anl.gov<br>> > > > To manage subscription options or unsubscribe:<br>> > > > https://lists.mcs.anl.gov/mailman/listinfo/mpich-discuss<br>> > ><br>> > > --<br>> > > Pavan Balaji<br>> > > http://www.mcs.anl.gov/~balaji<br>> <br>> -- <br>> Pavan Balaji<br>> http://www.mcs.anl.gov/~balaji<br></div> </div></body>
</html>