Within this project my task is the investigation of techniques for
creating a metacomputing environment. This includes
the specification of a software architecture, the communication protocol,
and the design and implemention of a secure framework with DCE integration.
In addition, I was involved in a project of establishing a DCE production
cell
for centralized managed systems. I gained detailed knowledge on the
DCE/Kerberos
interaction (Doug Engert, ANL) as well as on integrating DCE in a 3-tier
architecture. I also was responsible for specifying and
adopting a system management software for automatic system supervision.
Within another project which also included a beta-test for the new Sun
Enterprise Network Array I investigated the usage of high-end RAID-Arrays
for file servers. The performance measurements included tuning analysis for
different networking hardware (FDDI, OC-3, OC-12, Gigabit-Ethernet)
and system parameters.
I was responsible for developing a UNIX training course
[sl
ides].
Furthermore
I am giving a series of lectures for mathematic technical assistents
(assembler
programming, algorithms and datastructures).
13 years of programming practice mainly in the programming language C on
UNIX ( SunOS, Solaris, AIX, Unicos, Paragon OSF, ...), but also experience
in C++, Java, FORTRAN, PASCAL, Modula-2, PL/I, 370 Assembler and VAX/VMS.
I gained very detailed UNIX and networking knowledge while involved in
several
projects integrating heterogenous systems.
Since 1993 I am examiner for the study of mathematic technical assistent at
the industry and business foundation (IHK) of Cologne.
During my work at the HPCM project I established a testbed, using an
SSLeay patched Apache httpd, including DCE authentication for page access.
I also received detailed knowledge on programming the Kerberos software
in a DCE environment. I did some modifications to Doug Engerts (ANL)
k5dcelogin
for DCE 1.1.
I gained professional experience in presenting my work in German, as well as
in English. My paper "High-Performance Computer Management Based on Java"
was selected to be published in the upcomming
""Special
Issue HPCN'98 of the Future Generation Computer Systems". This issue
contains the
order of 12 to 15 of the best papers of the annual HPCN conference.
List of publications writen in English.
Hyperlinked titles are available at
Forschungszent
rum Jülich GmbH.
-
The VM/CMS Interface to the Common File System - Project-Documentation,
Technical Report 1990, IB-9019
-
Automatic Supervision of CRAY UNICOS Systems,
Proceedings of the CRAY User Gruoup Meeting (Spring), March 1991, London
pp. 279 - 284
-
Automatic Checkpointing of NQS Batch Jobs on CRAY UNICOS Systems,
Proceedings of the CRAY User Group Meeting(Spring), March 1993, Montreux,
pp. 250-255
-
Networks for Supercomputers at KFA-Juelich,
Proceedings of the CRAY User Group Meeting(Spring), March 1993, Montreux,
pp. 203-209
-
Operating the Data Migration Facility (DMF) at KFA,
Proceedings of the CRAY User Group Meeting(Spring), March 1993, Montreux,
pp. 121-124
-
A Metacomputer Architecture Based on Cooperative Resource Management,
Proceedings of High Performance Computing and Networking Europe 1997
(HPCN 1997), Wien, 28. - 30. April 1997
-
A Metacomputer Architecture Based on DCE,
Proceedings of DECORUM'97, Long Beach, California, 3 - 5 March, 1997
-
High-Performance Computer Management Based on Java,
Proceedings of High Performance Computing and Networking Europe 1998
(HPCN 1998), Amsterdam, 21. - 24. April 1998