Hello,
I am running into issues with my output when running Einstein Toolkit on our supercomputer.
I am using the intel 11.1 compiler, and in my options list when building the toolkit I have these options,
CC = icc CXX = icpc F77 = ifort F90 = ifort HDF5_DIR = BUILD
The simulation appears to run fine but the output is corrupt and I am receiving the following error message,
WARNING[L1,P5] (CarpetIOHDF5): HDF5 call 'attr = H5Acreate (dataset, "ioffset", H5T_NATIVE_INT, dataspace, H5P_DEFAULT)' returned error code -1 HDF5-DIAG: Error detected in HDF5 (1.8.6) thread 0: #000: H5A.c line 916 in H5Awrite(): not an attribute major: Invalid arguments to routine minor: Inappropriate type
Any suggestions on how to fix this error?
Tim,
I don't have an answer for you, but I have also seen this problem. For me, the problem does not occur when running a single-rank job. Is that true for you as well?
-Hal
On Wed, 2011-09-07 at 16:08 +0100, Tim Lemon wrote:
Hello,
I am running into issues with my output when running Einstein Toolkit on our supercomputer.
I am using the intel 11.1 compiler, and in my options list when building the toolkit I have these options,
CC = icc CXX = icpc F77 = ifort F90 = ifort HDF5_DIR = BUILD
The simulation appears to run fine but the output is corrupt and I am receiving the following error message,
WARNING[L1,P5] (CarpetIOHDF5): HDF5 call 'attr = H5Acreate (dataset, "ioffset", H5T_NATIVE_INT, dataspace, H5P_DEFAULT)' returned error code -1 HDF5-DIAG: Error detected in HDF5 (1.8.6) thread 0: #000: H5A.c line 916 in H5Awrite(): not an attribute major: Invalid arguments to routine minor: Inappropriate type
Any suggestions on how to fix this error?
Tim
We would need more information to debug this further. Could you open a ticket on trac.einsteintoolkit.org, and attach the following information:
- name and description of your supercomputer - complete option list - complete thorn list - content of the generated file configs/*/config-info - content of configs/CONFIG/config-data/make.config.defn - screen output when configuring Cactus (you may need to configure from scratch) - command you used to start your job - complete stdout/stderr when running - output of "ldd exe/cactus_CONFIG"
where CONFIG is the name you gave your configuration.
Also, what do you mean by "output is corrupt"? What are the symptoms?
-erik
On Wed, Sep 7, 2011 at 11:08 AM, Tim Lemon tim.lemon@soton.ac.uk wrote:
Hello,
I am running into issues with my output when running Einstein Toolkit on our supercomputer.
I am using the intel 11.1 compiler, and in my options list when building the toolkit I have these options,
CC = icc CXX = icpc F77 = ifort F90 = ifort HDF5_DIR = BUILD
The simulation appears to run fine but the output is corrupt and I am receiving the following error message,
WARNING[L1,P5] (CarpetIOHDF5): HDF5 call 'attr = H5Acreate (dataset, "ioffset", H5T_NATIVE_INT, dataspace, H5P_DEFAULT)' returned error code -1 HDF5-DIAG: Error detected in HDF5 (1.8.6) thread 0: #000: H5A.c line 916 in H5Awrite(): not an attribute major: Invalid arguments to routine minor: Inappropriate type
Any suggestions on how to fix this error?
-- Kind Regards
Tim lemon
Users mailing list Users@einsteintoolkit.org http://lists.einsteintoolkit.org/mailman/listinfo/users
users@lists.einsteintoolkit.org
-
Erik Schnetter -
Hal Finkel -
Tim Lemon