UCL-RITS / rcps-buildscripts

Scripts to automate package builds on RC Platforms
MIT License
39 stars 26 forks source link

Install request: Octopus 6 #77

Closed heatherkellyucl closed 7 years ago

heatherkellyucl commented 7 years ago

IN01814369

http://www.tddft.org/programs/octopus/wiki/index.php/Octopus_6.0

heatherkellyucl commented 7 years ago

Libxc 3.0.0 is recommended - we have 2.2.2

heatherkellyucl commented 7 years ago

Made a new script for installing Octopus 6.0 as the configure options have changed somewhat - there is no --enable-utils any more, and you need to give it --with-fftw-prefix and --with-netcdf-include (for the Fortran include).

heatherkellyucl commented 7 years ago

Failed some tests:

    Passed:  110 / 122
    Skipped: 11 / 122
    Failed:  1 / 122

    testfile                                                    # failed testcases
    ------------------------------------------------------------------------------
    periodic_systems/09-etsf_io.test                            4

Total run-time of the testsuite: 00:35:47

make[3]: *** [check-base] Error 1
make[3]: Leaving directory `/tmp/octopus-build.8Cz4NF15/octopus-6.0/testsuite'
make[2]: *** [check-wrapper] Error 2
make[2]: Leaving directory `/tmp/octopus-build.8Cz4NF15/octopus-6.0/testsuite'
make[1]: *** [check] Error 2
make[1]: Leaving directory `/tmp/octopus-build.8Cz4NF15/octopus-6.0/testsuite'
make: *** [check-recursive] Error 1
heatherkellyucl commented 7 years ago

This is the test output:

Starting test 98: periodic_systems/11-silicon_force.test
In execution: finite_systems_3d/02-curvilinear_coordinates.test functionals/04-oep.test linear_response/01-c
asida.test linear_response/05-polarizability.test optimal_control/04-box.test optimal_control/08-velocities.
test periodic_systems/04-silicon.test periodic_systems/06-h2o_pol_lr.test periodic_systems/07-tb09.test peri
odic_systems/08-benzene_supercell.test periodic_systems/09-etsf_io.test periodic_systems/11-silicon_force.te
st 

No mpiexec found: running in serial.

 ***** ETSF_IO *****  

Using workdir    : /tmp/octopus-.-periodic_systems-09-etsf_io.test.kt3Dai
Using test file  : ./periodic_systems/09-etsf_io.test 

Using input file : ./periodic_systems/09-etsf_io.01-gs.inp
Executing: cd /tmp/octopus-.-periodic_systems-09-etsf_io.test.kt3Dai; /bin/nice /tmp/octopus-build.8Cz4NF15/
octopus-6.0/testsuite/../src/main/octopus > out 

** Warning:
**   The value for 'Spacing' is inconsistent with the recommended value.
**   given    0.729 b > recommended    0.520 b

** Warning:
**   Wave-functions in Fourier space is under development.
**   It might not work or produce wrong results.

** Warning:
**   Some of the states are not fully converged!

        Elapsed time:     56.8 s

 Execution                              :        [   OK   ] 
 Total k-points                         :        [   OK   ]      (Calculated value = 2) 
 Total energy                           :        [   OK   ]      (Calculated value = -31.42700001) 
 Ion-ion energy                         :        [   OK   ]      (Calculated value = -31.43120280) 
 Eigenvalues sum                        :        [   OK   ]      (Calculated value = -0.47845186) 
 Hartree energy                         :        [   OK   ]      (Calculated value = 2.48290009) 
 Exchange energy                        :        [   OK   ]      (Calculated value = -8.24466712) 
 Correlation energy                     :        [   OK   ]      (Calculated value = -1.50771586) 
 Kinetic energy                         :        [   OK   ]      (Calculated value = 13.22648859) 
 External energy                        :        [   OK   ]      (Calculated value = -5.95279232) 
 k-point 1 (x)                          :        [   OK   ]      (Calculated value = 0.000000) 
 k-point 1 (y)                          :        [   OK   ]      (Calculated value = 0.000000) 
 k-point 1 (z)                          :        [   OK   ]      (Calculated value = 0.000000) 
 Eigenvalue  1                          :        [   OK   ]      (Calculated value = -0.280476) 
 Eigenvalue  8                          :        [   OK   ]      (Calculated value = 0.052868) 
 Eigenvalue 16                          :        [   OK   ]      (Calculated value = 0.160610) 
 Eigenvalue 18                          :        [   OK   ]      (Calculated value = 0.165239) 
 k-point 2 (x)                          :        [   OK   ]      (Calculated value = 0.500000) 
 k-point 2 (y)                          :        [   OK   ]      (Calculated value = 0.500000) 
 k-point 2 (z)                          :        [   OK   ]      (Calculated value = 0.500000) 
 Eigenvalue  1                          :        [   OK   ]      (Calculated value = -0.191934) 
 Eigenvalue  8                          :        [   OK   ]      (Calculated value = -0.098940) 
 Eigenvalue 16                          :        [   OK   ]      (Calculated value = 0.115142) 
 Eigenvalue 18                          :        [   OK   ]      (Calculated value = 0.217481) 

-----------------------------------------

Match density-etsf.nc  :

   Calculated value : 25776
   Reference value  : 25780
   Difference       : 4
   Tolerance        : 0.0001

 density-etsf.nc                        :        [  FAIL  ] 

-----------------------------------------

-----------------------------------------

Match geometry-etsf.nc :

   Calculated value : 3708
   Reference value  : 3712
   Difference       : 4
   Tolerance        : 0.0001

 geometry-etsf.nc                       :        [  FAIL  ] 

-----------------------------------------

-----------------------------------------

Match wfs-etsf.nc      :

   Calculated value : 1585316
   Reference value  : 1585320
   Difference       : 4
   Tolerance        : 0.0001

 wfs-etsf.nc                            :        [  FAIL  ] 

-----------------------------------------

-----------------------------------------

Match wfs-pw-etsf.nc   :

   Calculated value : 836064
   Reference value  : 836068
   Difference       : 4
   Tolerance        : 0.0001

 wfs-pw-etsf.nc                         :        [  FAIL  ] 

-----------------------------------------

 density.ncdf                           :        [   OK   ]      (Calculated value = 22248) 
 netcdf wfn 3                           :        [   OK   ]      (Calculated value = 44320) 
 netcdf wfn 4                           :        [   OK   ]      (Calculated value = 44320) 

Status: 4 failures

 ************************

The match is checking the binary size of the files...

testsuite/periodic_systems/09-etsf_io.test

# ETSF_IO files are binary, so we cannot easily test the contents, but
# at least we can check that this run finished without crashing,
# and test the size. However, these sizes are correct only for ETSF_IO 1.0.3;
# ETSF_IO 1.0.2 will give ones 208 smaller.

match ; density-etsf.nc  ; SIZE(static/density-etsf.nc)  ;   25780
match ; geometry-etsf.nc ; SIZE(static/geometry-etsf.nc) ;    3712
match ; wfs-etsf.nc      ; SIZE(static/wfs-etsf.nc)      ; 1585320
match ; wfs-pw-etsf.nc   ; SIZE(static/wfs-pw-etsf.nc)   ;  836068

match ; density.ncdf     ; SIZE(static/density.ncdf)        ;   22248
match ; netcdf wfn 3     ; SIZE(static/wf-k001-st0003.ncdf) ;   44320
match ; netcdf wfn 4     ; SIZE(static/wf-k001-st0004.ncdf) ;   44320

We are using ETSF_IO 1.0.4 (and were for the previous Octopus version). I guess we can ignore this.

heatherkellyucl commented 7 years ago

To do:

The first is currently building on Legion and Grace.

heatherkellyucl commented 7 years ago

Modules now available on Legion and Grace.

octopus/6.0/gnu-4.9.2
octopus/6.0-ompi/gnu-4.9.2
heatherkellyucl commented 7 years ago

This was waiting to see if the new version still has a problem where two consecutive Octopus runs in the same jobscript would fail, but work as two separate jobs.

* Could not read KS orbitals from 'restart/gs/
* Please run a ground-state calculation first!

(If not, may want to check env and the full contents of $TMPDIR and /dev/shm before and after the first Octopus run, in case it is leaving some variables set or lock files).

heatherkellyucl commented 7 years ago

Have heard nothing more about this problem, so closing until it shows up again.