Dear ET users' support team, I'm new to the Einstein Toolkit and I'm experiencing an issue with the installation of the 2020 ET release on the Marconi A3 machine at CINECA, Italy. I'm using the 2017 Intel family compilers and libraries, as required in the /Cactus/simfactory/mdb/optionlists/marconiA3.cfg file. The installation seems to end up successfully, but running the "HelloWorld" test simulation results in errors related to the fact that GLIBCXX_3.4.21, needed by /simulations/HelloWorld/SIMFACTORY/exe/cactus_sim, is not found. In fact, the command
strings /cineca/prod/opt/compilers/intel/pe-xe-2017/binary/inspector/lib64/libstdc++.so.6 | grep GLIBCXX
gives the following output:
GLIBCXX_3.4 GLIBCXX_3.4.1 GLIBCXX_3.4.2 GLIBCXX_3.4.3 GLIBCXX_3.4.4 GLIBCXX_3.4.5 GLIBCXX_3.4.6 GLIBCXX_3.4.7 GLIBCXX_3.4.8 GLIBCXX_3.4.9 GLIBCXX_3.4.10 GLIBCXX_3.4.11 GLIBCXX_3.4.12 GLIBCXX_3.4.13 GLIBCXX_3.4.14 GLIBCXX_3.4.15 GLIBCXX_3.4.16 GLIBCXX_3.4.17 GLIBCXX_3.4.18 GLIBCXX_3.4.19 GLIBCXX_3.4.20 GLIBCXX_FORCE_NEW GLIBCXX_DEBUG_MESSAGE_LENGTH
'GLIBCXX_3.4.21' does not appear. This version of GLIBCXX is only available within the 2019 and 2020 Intel families, but these are not compatible with many of the libraries needed by the ET and available on Marconi A3. What can I do to get the ET to work?
The modules I have loaded on Marconi A3 are the following:
1) profile/base 2) profile/advanced 3) profile/archive 4) gnu/6.1.0 5) cmake/3.12.0 6) intel/pe-xe-2017--binary 7) intelmpi/2017--binary 8) zlib/1.2.8--gnu--6.1.0 9) szip/2.1--gnu--6.1.0 10) hdf5/1.8.17--intel--pe-xe-2017--binary 11) papi/5.7.1 12) git/2.17 13) hwloc/1.11.3--gnu--6.1.0 14) mkl/2017--binary 15) blas/3.6.0--intel--pe-xe-2017--binary 16) lapack/3.6.1--intel--pe-xe-2017--binary 17) gsl/2.2.1--intel--pe-xe-2017--binary 18) fftw/3.3.4--intelmpi--2017--binary 19) boost/1.61.0--intelmpi--2017--binary 20) python/3.6.4
The config-info file and a snapshot of the error messages generated by running the "HelloWorld" simulation are attached to the e-mail. Thanks in advance for your support. Best regards,
Lorenzo Ennoggi
