Crash calculation

[andre-cloud]

Hello

Is there a reason why Censo crashes the calculation during any part? It crashed on part0, 1, 2 and 3. I've changed only the functional, using one of the selected ones present in censo -tutorial.

In the output file it didn't say anything, but from the top there aren't orca processes running.

----------------------------------------------------------------------------------------------------
                                      CRE OPTIMIZATION - PART2                                      
----------------------------------------------------------------------------------------------------

program:                                                       orca
functional for part2:                                          wb97x-d4
basis set for part2:                                           3-21g
using the xTB-optimizer for optimization:                      on
using the new ensemble optimizer:                              on
optimize all conformers below this G_thr(opt,2) threshold:     2.5
Spearman threshold:                                            0.918
number of optimization iterations:                             8
optimization level in part2:                                   lax
solvent:                                                       hexane
solvent model applied in the optimization:                     smd
solvent model for Gsolv contribution:                          cpcm
temperature:                                                   298.15
evalulate at different temperatures:                           on
temperature range:                                             273.15, 278.15, 283.15, 288.15, ...
Boltzmann sum threshold G_thr(2) for sorting in part2:         99.0
calculate mRRHO contribution:                                  on
program for mRRHO contribution:                                xtb
GFN version for mRRHO and/or GBSA_Gsolv:                       gfn2
Apply constraint to input geometry during mRRHO calculation:   on

Optimizing geometries at DFT level with implicit solvation!
The optimization is calculated for:
 CONF1,  CONF2,  CONF3,  CONF4,  CONF5,  CONF6,  CONF7,  CONF8,  CONF9, CONF10, CONF11
CONF12, CONF13, CONF14, CONF15, CONF16, CONF17, CONF18, CONF19, CONF20, CONF21, CONF22
CONF23, CONF24, CONF25, CONF26, CONF27, CONF28, CONF29, CONF30, CONF31, CONF32, CONF33
CONF34, CONF35, CONF36, CONF37

Constructed folders!

Preparing 37 calculations.
Tasks completed!

************************Starting optimizations************************

Starting threshold is set to 2.5 + 60.0 % = 4.0 kcal/mol

Lower limit is set to G_thr(opt,2) = 2.5 kcal/mol

*******************************CYCLE 1********************************

Starting 37 optimizations.
Running optimization in CONF1/wb97x-d4    
Running optimization in CONF2/wb97x-d4    
Running optimization in CONF3/wb97x-d4    
Running optimization in CONF4/wb97x-d4    
Running optimization in CONF5/wb97x-d4    
Running optimization in CONF6/wb97x-d4    
Running optimization in CONF7/wb97x-d4    
Running optimization in CONF8/wb97x-d4    
Running optimization in CONF9/wb97x-d4    
Running optimization in CONF10/wb97x-d4   
Running optimization in CONF11/wb97x-d4   
Running optimization in CONF12/wb97x-d4   
Running optimization in CONF13/wb97x-d4   
Running optimization in CONF14/wb97x-d4   
Running optimization in CONF15/wb97x-d4   
Running optimization in CONF16/wb97x-d4   
Running optimization in CONF17/wb97x-d4   
Running optimization in CONF18/wb97x-d4   
Running optimization in CONF19/wb97x-d4   
Running optimization in CONF20/wb97x-d4   
Running optimization in CONF21/wb97x-d4   
Running optimization in CONF22/wb97x-d4

This is the last file I have of a calculation that has crashed.

In addition to that, I have to say that on the same server there are two censo calculations running, but it crashes even when only one is running. How can I fix this problem (without using censo --restart which surely solve partially the problem)?

[fabothch]

Hello,

without information about what caused the crash, I can only guess. During the optimizations, the CONFX/wb97x-d4/opt-part2.out files are written which contain the output from xtb and orca. You could look at these files and maybe there is an error code from ORCA that can help diagnose what is going on. Maybe it has something to do with the basis set or the functional, but I am just guessing here.

[andre-cloud]

In every directory of CONF* of this specific calculation the less opt-part2.out | tail -1 is

CONF10/wb97x-d4/opt-part2.out:normal termination of xtb
CONF11/wb97x-d4/opt-part2.out:normal termination of xtb
CONF12/wb97x-d4/opt-part2.out:normal termination of xtb
CONF13/wb97x-d4/opt-part2.out:normal termination of xtb
CONF14/wb97x-d4/opt-part2.out:normal termination of xtb
CONF15/wb97x-d4/opt-part2.out:normal termination of xtb
CONF16/wb97x-d4/opt-part2.out:normal termination of xtb
CONF17/wb97x-d4/opt-part2.out:normal termination of xtb
CONF18/wb97x-d4/opt-part2.out:normal termination of xtb
CONF19/wb97x-d4/opt-part2.out:normal termination of xtb
CONF1/wb97x-d4/opt-part2.out:normal termination of xtb
CONF20/wb97x-d4/opt-part2.out:normal termination of xtb
CONF21/wb97x-d4/opt-part2.out:normal termination of xtb
CONF22/wb97x-d4/opt-part2.out:normal termination of xtb
CONF2/wb97x-d4/opt-part2.out:normal termination of xtb
CONF3/wb97x-d4/opt-part2.out:normal termination of xtb
CONF4/wb97x-d4/opt-part2.out:normal termination of xtb
CONF5/wb97x-d4/opt-part2.out:normal termination of xtb
CONF6/wb97x-d4/opt-part2.out:normal termination of xtb
CONF7/wb97x-d4/opt-part2.out:normal termination of xtb
CONF8/wb97x-d4/opt-part2.out:normal termination of xtb
CONF9/wb97x-d4/opt-part2.out:normal termination of xtb

[fabothch]

Yes, xtb was terminated normally, but ORCA might have crashed. xtb is only used as a driver here and calls ORCA. Please check the whole output and see if ORCA has finished normally.

[andre-cloud]

Yes, all Orca calculations report the **** ORCA TERMINATED NORMALLY **** statement

[fabothch]

What I can offer is: you pack the calculation folder as a zip archive and I can have a look at it. Without a hint of what exactly crashed, I can not diagnose anything.

[andre-cloud]

I will, thanks very much This calculation has to be restarted with the command --restart, unfortunately. Next time I surelly will