I don't think we have such an environment variable -- are you sure you're using
Open MPI?
The only reference to MPI_TYPE_MAX I see in the OMPI source tree is in the
ROMIO README file:
-
* When using ROMIO with SGI MPI, you may sometimes get an error
message from SGI MPI: ``MPI has run out o
Hi All,
I keep getting an error about running out of MPI_TYPE_MAX and needing to set
the environment variable higher. What is this, and why is it happening? All of
the types and groups, etc that I create during my programs run are freed at the
appropriate times. Making this number 10x bigger ge
Hi Jeff,
You were right. One of the other admins of the server I am working on, had a
script that runs the firewall every time I logged in. So even when I was
turning it off manually, the firewall ran the next time i logged in and hence
the error.
Thanks.
- Rohan
--- On Tue, 4/6/10, Jeff S
Hello,
I've noticed that ompi-restart doesn't support the --rankfile option.
It only supports --hostfile/--machinefile. Is there any reason
--rankfile isn't supported?
Suppose you have a cluster without a shared file system. When one node
fails, you transfer its checkpoint to a spare node and in
On Thu, Apr 8, 2010 at 10:31 AM, Jeff Squyres wrote:
> Yes. There is usually a difference between interactive logins and
> non-interactive logins on which paths, etc. get set. Look in your shell
> startup and see if there is somewhere that it exits early (or otherwise
> doesn't process) for n
Yes. There is usually a difference between interactive logins and
non-interactive logins on which paths, etc. get set. Look in your shell
startup and see if there is somewhere that it exits early (or otherwise doesn't
process) for non-interactive logins.
In short: you need to ensure that your
Dear OpenMPI users
We built OpenMPI 1.4.1 on a new cluster and get the following error message
when starting a test job from the master node:
ham4#mpirun -np 4 --host cn001 /path/hello
orted: error while loading shared libraries: libimf.so: cannot open shared
object file: No such file or dire
1. Ralph, if I try to do my experiment with SGE, then it's the same
results with 1.4.2 as with 1.4.1.
$ qrsh -cwd -V -pe ompi* 16 -l h_rt=10:00:00,h_vmem=2G bash
graphics03 $ cat hosts
graphics01 slots=1
graphics02 slots=1
graphics03 slots=1
graphics04 slots=1
graphics03 $ ~/openmpi/gnu129/bin
Serge writes:
> This is exactly what I am doing -- controlling distribution of
> processes with mpirun on the SGE-allocated nodes, by supplying the
> hostfile. Grid Engine allocates nodes and generates a hostfile, which
> I then can modify however I want to, before running the mpirun
> command. M
