NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.

To learn about what's going on, see About the Archive Site.

Re: [netcdfgroup] Issues with HDF5 1.8.10 and NetCDF 4.2.1.1: tst_parallel

  • To: "Thompson, Matt (GSFC-610.1)[SCIENCE SYSTEMS AND APPLICATIONS INC]" <matthew.thompson@xxxxxxxx>
  • Subject: Re: [netcdfgroup] Issues with HDF5 1.8.10 and NetCDF 4.2.1.1: tst_parallel
  • From: Nick Papior Andersen <nickpapior@xxxxxxxxx>
  • Date: Wed, 2 Jan 2013 21:14:19 +0100
I have the exact same problem, I am currently trying to figure out exactly
where it happens. But surely no update should be used. Keep the HDF5 1.8.9.
Maybe this should be added in the README file for a quick user guidance.

Kind regards Nick

2013/1/2 Thompson, Matt (GSFC-610.1)[SCIENCE SYSTEMS AND APPLICATIONS INC] <
matthew.thompson@xxxxxxxx>

> Folks,
>
> I was wondering if you could help me out with a minor issue. I work on the
> GEOS-5 GCM model at NASA Goddard and our Base Libraries include HDF5 and
> NetCDF4 as we are planning to transition to using parallel NetCDF4 soon in
> the future.
>
> At the moment, we are doing just fine testing/running with HDF5 1.8.9 and
> NetCDF 4.2.1.1, but I recently saw that HDF5 updated to 1.8.10 and I
> thought I'd test it in the interest of keeping up with current at least in
> my mind (I've gotten behind on libraries before). Plus, the INSTALL file
> does say: "Versions required are at least HDF5 1.8.8..."
>
> However, I find that while NetCDF compiles just fine with HDF5 1.8.10, the
> "make check" tests are another matter. In particular, tst_parallel "hangs":
>
> Testing very simple parallel I/O with 4 processors...
>
> *** tst_parallel testing very basic parallel access.
> *** tst_parallel testing whether we can create file for parallel access
> and write to it...
>
> This hang happens with both Intel 11.0.083 + Intel MPI 3.2.2.006 and Intel
> 13.0.0.079 + MVAPICH2 1.9a, so I don't think it's MPI-specific. Plus, those
> same combinations do "make check" just fine with HDF5 1.8.9+NetCDF 4.2.1.1.
>
> I looked at tst_parallel.c, a pretty simple program, and I couldn't see
> anything obvious so it's probably a bit more low-level than I can figure
> out at the moment.
>
> Obviously, my "solution" at present is "don't upgrade HDF5 at the moment",
> but I was wondering if, perhaps, there is something simple I might be
> missing from a configure or make step that should be there but isn't and
> was only just exposed by HDF5 1.8.10.
>
> Has anyone else encountered this issue?
>
> Thanks,
> Matt Thompson
>
> --
> Matt Thompson, PhD     SSAI, Sr Software Test Engr
> NASA GSFC, Global Modeling and Assimilation Office
> Code 610.1, 8800 Greenbelt Rd, Greenbelt, MD 20771
> Phone: 301-614-6712              Fax: 301-614-6246
>
> _______________________________________________
> netcdfgroup mailing list
> netcdfgroup@xxxxxxxxxxxxxxxx
> For list information or to unsubscribe,  visit:
> http://www.unidata.ucar.edu/mailing_lists/
  • 2013 messages navigation, sorted by:
    1. Thread
    2. Subject
    3. Author
    4. Date
    5. ↑ Table Of Contents
  • Search the netcdfgroup archives: