[molpro-user] Test job h2o_select hanging dual-Xeon Fedora2 node

[Dr Seth OLSEN]

Hello Molpro-Users,

I am in the process of installing molpro on a dual Xeon SunFire v60x node running Fedora Core 2.  I 've been trying to run the test suite and I have a problem.  The test job h2o_select does not end.  Furthermore, the molpro.exe excutable is unkillable.  It gets as far as the integral sort and then writes no more to output, though it keeps running at ~99%CPU.  I notice that periodically it outputs errors such as:
 ?WARNING: TOO HIGH ADDRESS IN CORLSI: LT=********  IBASE=********  LTOP= -46216617  MEMSTACK= -42244435 

I'm not sure what this means.  I cannot help but notice that the previous job h2o_vdz, which does finish prints out a similar warning:

 ?WARNING: TOO HIGH ADDRESS IN CORLSI: LT=********  IBASE=********  LTOP= -42770350  MEMSTACK= -42301715

It does appear that MolPro is creating files in the designated scratch space.  An 'ls -l' in this directory gives the following output:
-rw-r--r--  1 root root    0 Nov 12 11:33 7584
-rw-------  1 root root    0 Nov 12 11:33 fort62j9YF
-rw-------  1 root root   69 Nov 12 11:07 fortAU6rMx
-rw-------  1 root root   69 Nov 12 11:33 fortAXMdCh
-rw-------  1 root root 1967 Nov 12 11:07 fortdxbBl4
-rw-------  1 root root  402 Nov 12 11:07 fortgr7bEu
-rw-------  1 root root    0 Nov 12 11:07 fortpkzjd1
-rw-------  1 root root 1967 Nov 12 11:33 fortUI3ifT
-rw-------  1 root root  402 Nov 12 11:33 fortYFo6l4

I previously had no problem running the test suite on a single cpu P4 machine, also running Fedora Core 2.  

Does anybody have any insight into the cause of the problem?

Cheers,

Seth Olsen





ccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccms

Dr Seth Olsen, PhD
Postdoctoral Fellow, Computational Systems Biology Group
Centre for Computational Molecular Science
Chemistry Building,
The University of Queensland
Qld 4072, Brisbane, Australia

tel (617) 33653732
fax (617) 33654623
email: s.olsen1 at uq.edu.au
Web: www.ccms.uq.edu.au 

ccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccms