CCSD(T) job error message
Clicked A Few Times
6:37:22 AM PDT - Mon, Oct 6th 2014
Hello,

I got error messages of running big CCSD(T) jobs.

 This molecule is running with aQ basis set, totally 1500 basis functions and 48 correlated orbitals (28 in the core).


1) With version of 6.3, at the step of CCSD,

the error shows "eaf_write: rc ne bytes -1999 bytes 295714328
                                 ccsd_sxy: write failed        5"

I requested 16 cores per node on CASCADE.

Does that mean I need to use less cores per node ?



2) Also I tried the version of 6.5, also at the CCSD step,
got the error message :


000: g_st2 size: 19855 MB
000: mem. avail 20698 MB
000: Not enough mem for st2. Increase MA by 3296
000: MB
000: Not Enough Memory to keep ST2 in local mem


I already put lots of stack memory, around 20000mb, do i need to put more ?
I thik it is too much. There maybe something wrong.
Is it necessary to put more global memory at the same time ?




Thanks a lot.


Zongtang

Forum Vet
12:32:25 PM PDT - Mon, Oct 6th 2014
Zontang,
My advice is that, for systems similar to cascade, you should be using eight (or even four) processes on each node. This allows you both to have more memory available for each process and you would get better communication performance because fewer processes are sharing the same network interface.

"ccsd_sxy: write failed" failure: you are probably filling up the /scratch filesystem on cascade. My suggestion is to use the direct approach ("nodisk" keywork in the CCSD input section)

"g_st2" warnings: the warning message is advising you to add another 3.3GB of memory of MA a.k.a. stack memory.
I strongly encourage you to do that since the parallel performance improves dramatically. You might have to use just 4 cores/node to accommodate this memory request.

Clicked A Few Times
12:43:09 PM PDT - Mon, Oct 6th 2014
Dear Edo,

Thanks a lot.

I will try.


Forum >> NWChem's corner >> Running NWChem