CCL:G: problem wirh g03 D.02

From: "alberto savoini" <alberto.savoini],[polimerieuropa.com>
Subject: CCL:G: problem wirh g03 D.02
Date: Mon, 6 Nov 2006 11:07:55 -0500

 Sent to CCL by: "alberto  savoini"
 [alberto.savoini(a)polimerieuropa.com]
 I've installed  the upgrade for Gaussian ( g03 D.02  s/n FC8441343010) and LINDA
 ( Rev 7.1 s/n FC8341343020 ).
 My cluster has  fedora  core 3 and PGI 6.1.
 While the compilation is OK I have a problem running parallel jobs
 (%NProcShared=2 %NProcLinda=2)  :
 setenv GAUSS_EXEDIR
 /opt2/g03d02/g03/linda-exe:/opt2/g03d02/g03/bsd:/opt2/g03d02/g03/local:/opt2/g03d02/g03/extras:/opt2/g03d02/g03
 g03 test004.com
 Incorrectly built binary which accesses errno or h_errno directly. Needs to be
 fixed.
 ntsnet: using executable file /opt2/g03d02/g03/linda-exe/l302.exel
 ntsnet: trying to schedule 1 worker
 ntsnet: scheduled a total of 1 worker
 ntsnet: starting master process on node11.cluster
 ntsnet: starting 1 worker on node12.cluster
 l302.exel: machine.c:365: _l_md_post_local_recv: Assertion `!local_recv_sanity'
 failed.
 eval server 0 on node11.cluster has dropped it's connection.
 If I run a job using only one node ( 2 CPU ) it works correctly.
 Alberto Savoini - Sistemi Informativi
 Eni S.p.A. - Divisione Refining & Marketing
 Centro Ricerche di Novara - Istituto Guido Donegani
 Via Fauser,4 28100 Novara (NO) - Italy
 Tel. +39 0321 447505
 Fax  +39 0321 447679
 email: alberto.savoini=polimerieuropa.com