Open MPI logo

Open MPI User's Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Open MPI User's mailing list

Subject: Re: [OMPI users] mpi_init waits 64 seconds if vpn is connected
From: Ralph Castain (rhc_at_[hidden])
Date: 2013-03-22 10:14:19


Ah, yes - that pinpoints the problem. It looks like your vpn is indeed interfering with connections. I suspect it must be some kind of vpn/Ubuntu configuration issue as I have the same experience on my Mac laptop that you report (i.e., no issue).

I'd suggest googling the Ubuntu site (or just in general) to see if someone has a suggestion. Afraid I'm not a Ubuntu nor vpn person and don't have an immediate answer. :-(

FWIW: it isn't a specific connection - we just ask for any port and use the standard Linux "connect" function.

On Mar 22, 2013, at 5:12 AM, David A. Boger <dab143_at_[hidden]> wrote:

> Following up on your TCP remark, I found that during the delay, netstat -tnp shows
>
> tcp 0 1 192.168.1.3:56343 192.168.1.3:47830 SYN_SENT 24191/mpi_hello
>
> and that while the vpn is connected, I am unable to ping 192.168.1.3 (the machine I am on).
>
> On the other hand, on the laptop (where mpi runs fine with vpn connected), the ping-to-self works fine even when the vpn is connected.
>
>
>
>
> On Fri, Mar 22, 2013 07:37 AM, "David A. Boger" <dab143_at_[hidden]> wrote:
> Thanks Ralph. I have a Mac OS X 10.6.8 laptop where I can run open-mpi 1.2.8 and open-mpi 1.6.4 with the vpn connected without any problem, even without having to exclude the vpn interface, so you're probably right -- the existence of the vpn interface alone doesn't explain the problem. Nevertheless, disconnecting the vpn on my ubuntu box definitely resolves the problem, so I think it's tied in somehow.
>
> Do you think the process is hanging looking for a specific TCP connection, or just any TCP connection? If it's a specific one, is there a way to find out which or to test using something outside of mpirun that would show the same delay?
>
> Thanks again,
> David
>
> On Fri, Mar 22, 2013 12:25 AM, Ralph Castain <rhc_at_[hidden]> wrote:
> The process is hanging trying to open a TCP connection back to mpirun. I would
> have thought that excluding the vpn interface would help, but it could be that
> there is still some interference from the vpn software itself - as you probably
> know, vpn generally tries to restrict connections.
>
> I don't recall seeing this behavior with my laptop (which also runs with a
> Cisco vpn), but I'll check it again in the morning and let you know.
>
>
> On Mar 21, 2013, at 6:52 PM, David A. Boger <dab143_at_[hidden]> wrote:
>
> > I am having a problem on my linux desktop where mpi_init hangs for
> approximately 64 seconds if I have my vpn client connected but runs immediately
> if I disconnect the vpn. I've picked through the FAQ and Google but have failed
> to come up with a solution.
> >
> > Some potentially relevant information: I am using Open MPI 1.4.3 under
> ubuntu 12.04.1 and Cisco AnyConnect VPN Client. (I have also downloaded
> openmpi 1.6.4 and built it from source but believe it behaves the same
> way.)
> >
> > Some potentially irrelevant information: I believe SSH tunneling is
> disabled by the vpn. While the vpn is connected, ifconfig shows an extra
> interface (cscotun0 with inet addr:10.248.17.27 that shows up in the
> contact.txt file:
> >
> > wt217:~/wrk/mpi> cat
> /tmp/openmpi-sessions-dab143_at_wt217_0/29142/contact.txt
> > 1909850112.0;tcp://192.168.1.3:48237;tcp://10.248.17.27:48237
> > 22001
> >
> > The code is simply
> >
> > #include <stdio.h>
> > #include <mpi.h>
> >
> > int main(int argc, char** argv)
> > {
> > MPI_Init(&argc, &argv);
> > MPI_Finalize();
> > return 0;
> > }
> >
> > I compile it using "mpicc -g mpi_hello.c -o mpi_hello" and
> execute it using "mpirun -d -v ./mpi_hello". (The problem occurs
> whether or not I asked for more than one processor.) With verbosity on, I
> get the following output:
> >
> > wt217:~/wrk/mpi> mpirun -d -v ./mpi_hello
> > [wt217:22015] procdir: /tmp/openmpi-sessions-dab143_at_wt217_0/29144/0/0
> > [wt217:22015] jobdir: /tmp/openmpi-sessions-dab143_at_wt217_0/29144/0
> > [wt217:22015] top: openmpi-sessions-dab143_at_wt217_0
> > [wt217:22015] tmp: /tmp
> > [wt217:22015] [[29144,0],0] node[0].name wt217 daemon 0 arch ffc91200
> > [wt217:22015] Info: ! Setting up debugger process table for applications
> > MPIR_being_debugged = 0
> > MPIR_debug_state = 1
> > MPIR_partial_attach_ok = 1
> > MPIR_i_am_starter = 0
> > MPIR_proctable_size = 1
> > MPIR_proctable:
> > (i, host, exe, pid) = (0, wt217,
> /home/dab143/wrk/mpi/./mpi_hello, 22016)
> > [wt217:22016] procdir: /tmp/openmpi-sessions-dab143_at_wt217_0/29144/1/0
> > [wt217:22016] jobdir: /tmp/openmpi-sessions-dab143_at_wt217_0/29144/1
> > [wt217:22016] top: openmpi-sessions-dab143_at_wt217_0
> > [wt217:22016] tmp: /tmp
> > <hangs for approximately 64 seconds>
> > [wt217:22016] [[29144,1],0] node[0].name wt217 daemon 0 arch ffc91200
> > [wt217:22016] sess_dir_finalize: proc session dir not empty - leaving
> > [wt217:22015] sess_dir_finalize: proc session dir not empty - leaving
> > [wt217:22015] sess_dir_finalize: job session dir not empty - leaving
> > [wt217:22015] sess_dir_finalize: proc session dir not empty - leaving
> > orterun: e! xiting with status 0
> >
> > The code hangs for approximately 6! 4 second s after the line that reads
> "tmp: /tmp".
> >
> > If I attach gdb to the process during this time, the stack trace
> (attached) shows that the pause is in __GI___poll in
> /sysdeps/unix/sysv/linux/poll.c:83.
> >
> > If I add "-mca oob_tcp_if_exclude cscotun0", then the
> corresponding address for that vpn interface no longer shows up in contact.txt,
> but the problem remains. I also add "-mca btl ^cscotun0 -mca
> btl_tcp_if_exclude cscotun0" with no effect.
> >
> > Any idea what is hanging this up or how I can get more information as to
> what is going on during the pause? I assume connecting the vpn has caused
> mpi_init to look for something that isn't available and that eventually times
> out, but I don't know what.
> >
> > Output from ompi_info and the gdb stack trace is attached.
> >
> > Thanks,
> > David
> >
> >
> <stack.txt.bz2><ompi_info.txt.bz2>_______________________________________________
> > users mailing list
> > users_at_[hidden]
> > http://www.open-mpi.org/mailman/listinfo.cgi/users
>
>
> _______________________________________________
> users mailing list
> users_at_[hidden]
> http://www.open-mpi.org/mailman/listinfo.cgi/users
>
>
>
>
>
> _______________________________________________
> users mailing list
> users_at_[hidden]
> http://www.open-mpi.org/mailman/listinfo.cgi/users
>
>
>
> _______________________________________________
> users mailing list
> users_at_[hidden]
> http://www.open-mpi.org/mailman/listinfo.cgi/users