- aktuell - lists.lrz.de

GPFS problems on SuperMUC
by LRZ aktuell Oct. 12, 2012

Oct. 12, 2012

Dear users of SuperMUC, we currently experience problems with GPFS (filesystems $WORK and $SCRATCH) on SuperMUC. Access to most of the data is still possilbe, but some files or directories are no longer accessible. IBM support has been notified and is working on the problem. We apologize for the inconvenience caused by this failure. This information is also available on our web server http://www.lrz-muenchen.de/services/compute/supermuc/aktuell/ali4430/ Hans-Georg Kleinhenz

1 0

Upd:Linux Cluster: Serial Segment Maintenance (Update)
by LRZ aktuell Oct. 11, 2012

Oct. 11, 2012

Aenderungen an der Mitteilung: Prolongation of maintenance Update (October 11, 17:30) Unfortunately there still are issues with the SLURM configuration of the serial cluster parts, so we need to prolong the maintenance until these problems have been resolved. We'll keep you updated on the progress via this document. ----------------------------------------------------------------------- Dear users of the Linux Cluster Systems at LRZ, The overdue operating system update of the serial segment has been scheduled for October 8, 8:00 am - October 11, 6:00 pm. At the beginning of this maintenance, all SGE jobs that are still running will be deleted. All queued jobs that target the systems to be upgraded will be lost. During the above interval, the cluster will not be available for user operation. After return to user operation, you will need to change your SGE job scripts over to using SLURM; documentation on how to use SLURM for serial jobs will be published on the LRZ web server and linked from here as it becomes available. Generally, it may also be necessary for you to re-compile your applications for the new environment. Also note that access to the cluster login node lx64ia2 will already become unavailable on October 2; please use the remaining login node lx64ia3 in the meantime. The secure shell keys for all login nodes may become invalid and will be updated on the LRZ web server if that is the case. Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/compute/aktuell/ali4423/ Reinhold Bader

1 0

Upd:MPP Cluster: Returned to user operation
by LRZ aktuell Oct. 11, 2012

Oct. 11, 2012

Changes to this message: Update The MPP cluster has been returned to user operation after a rather over-extended maintenance; apart from difficulties with integrating the new hardware there also was some fallout from the recent power interruption which required more extended repairs than expected. This information is also available on our web server http://www.lrz-muenchen.de/services/compute/aktuell/ali4403/ Reinhold Bader

1 0

Upd:TSM-Serverprobleme am 10.10.2012
by LRZ aktuell Oct. 10, 2012

Oct. 10, 2012

Aenderungen an der Mitteilung: Update UPDATE 10.10.2012 17:00 Uhr Alle Server sind wieder verfuegbar. Allerdings mit eingeschraenkter Performance auf S12 - S33 Nach dem letzten Betriebssystemupdate scheint sich ein Problem mit den 10GE Netzwerktreibern auf den TSM Servern S12-S33 eingeschlichen zu haben. Diese treten anscheinend nur unter hoher Last auf, weshalb es in der Testumgebung nicht aufgetreten ist. Als Workaround sind die o.g. Server nun mit 1GE ans Netz angebunden. Es kann also zu laenger laufenden Backups als gewohnt kommen. Wir entschuldigen uns fuer die Unannehmlichkeiten. Orginalnachricht: Auf Grund einiger Probleme, sind derzeit die TSM - Server nicht verfuegbar. An der Beseitigung der Stoerung wird gearbeitet. Wir bitten um Ihr Verstaendnis. Ihr ABS-Team Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/datenhaltung/adsm/aktuell/ali4429/ Monika Hufnagl

1 0

TSM-Serverprobleme am 10.10.2012
by LRZ aktuell Oct. 10, 2012

Oct. 10, 2012

Auf Grund einiger Probleme, sind derzeit die TSM - Server nicht verfuegbar. An der Beseitigung der Stoerung wird gearbeitet. Wir bitten um Ihr Verstaendnis. Ihr ABS-Team Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/datenhaltung/adsm/aktuell/ali4429/ Monika Hufnagl

1 0

Upd:Update: 10th Tuning Workshop of the Virtual Institute - High Productivity Supercomputing (VI-HPS)
by LRZ aktuell Oct. 10, 2012

Oct. 10, 2012

Changes to this message: Update Please accept our apologies if you encountered problems registering for the 10th VI-HPS Tuning Workshop which will take place at LRZ in Garching next week, 16-19 October 2012. Places are still available, and registration has been extended to Friday 12th October. Agenda and further information at http://www.vi-hps.org/vi-hps-tw10/ "The 9th VI-HPS Tuning Workshop was quite a boost for me in my work to optimize [our] main application, which is a hybrid MPI + OpenMP parallel code." -- previous workshop participant Original message: ********************************************************************* 10th Tuning Workshop of the Virtual Institute - High Productivity Supercomputing (VI-HPS) at Leibniz Rechenzentrum (LRZ), Garching-bei-Muenchen, Germany 16-19 October 2012 http:/www.vi-hps.org/ vi-hps-tw10/ ********************************************************** *********** To help improve the quality and accelerate the development process of complex simulation codes in science and engineering, the VI-HPS Virtual Institute - High Productivity Supercomputing develops tools for checking the correctness and analyzing the performance of parallel applications. Emphasis is given to the programming models MPI & OpenMP. This workshop organized by VI-HPS for Gauss Centre for Supercomputing PRACE Advanced Training Centre will: * give an overview of our programming tools suite * explain the functionality of individual tools, and how to use them effectively * offer hands-on experience and expert assistance using the tools Tools which will be covered include Kcachegrind, MAQAO, PAPI, Periscope, Scalasca, TAU, Vampir and Score-P. The workshop will be held in English and runs from 09:00 to not later than 18:00 each day, with breaks for lunch and refreshments. There is no fee for participation, however, participants are responsible for their own travel and accommodation arrangements. Classroom capacity is limited, therefore priority will be given to applicants with parallel codes already running on the LRZ IBM SuperMUC & SuperMIG, and those bringing codes from similar systems to work on. Participants are therefore strongly encouraged to prepare their own MPI, OpenMP and hybrid MPI+OpenMP parallel application codes for analysis. For further information, please visit the workshop web page at: http:// www.vi-hps.org/vi-hps-tw10/ For information about other VI-HPS workshops and tutorials, and to sign up to the mailing list for announcements, visit: http://www.vi-hps.org/ training/ Contact: Ventsislav Petkov (TUM) +49 89 289-17680 Brian Wylie (JSC) +49 2461 61-6589 ********************************************************************* Visit www.vi-hps.org for more information about the Virtual Institute - High Productivity Supercomputing. ************************************* ******************************** This information is also available on our web server http://www.lrz-muenchen.de/services/compute/aktuell/ali4410/ Wolfram Hesse

1 0

Upd:Update: 10th Tuning Workshop of the Virtual Institute - High Productivity Supercomputing (VI-HPS)
by LRZ aktuell Oct. 10, 2012

Oct. 10, 2012

Changes to this message: Update Please accept our apologies if you encountered problems registering for the 10th VI-HPS Tuning Workshop which will take place at LRZ in Garching next week, 16-19 October 2012. Places are still available, and registration has been extended to Friday 12th October. Agenda and further information at http://www.vi-hps.org/vi-hps-tw10/ "The 9th VI-HPS Tuning Workshop was quite a boost for me in my work to optimize [our] main application, which is a hybrid MPI + OpenMP parallel code." -- previous workshop participant Original message: ********************************************************************* 10th Tuning Workshop of the Virtual Institute - High Productivity Supercomputing (VI-HPS) at Leibniz Rechenzentrum (LRZ), Garching-bei-Muenchen, Germany 16-19 October 2012 http:/www.vi-hps.org/ vi-hps-tw10/ ********************************************************** *********** To help improve the quality and accelerate the development process of complex simulation codes in science and engineering, the VI-HPS Virtual Institute - High Productivity Supercomputing develops tools for checking the correctness and analyzing the performance of parallel applications. Emphasis is given to the programming models MPI & OpenMP. This workshop organized by VI-HPS for Gauss Centre for Supercomputing PRACE Advanced Training Centre will: * give an overview of our programming tools suite * explain the functionality of individual tools, and how to use them effectively * offer hands-on experience and expert assistance using the tools Tools which will be covered include Kcachegrind, MAQAO, PAPI, Periscope, Scalasca, TAU, Vampir and Score-P. The workshop will be held in English and runs from 09:00 to not later than 18:00 each day, with breaks for lunch and refreshments. There is no fee for participation, however, participants are responsible for their own travel and accommodation arrangements. Classroom capacity is limited, therefore priority will be given to applicants with parallel codes already running on the LRZ IBM SuperMUC & SuperMIG, and those bringing codes from similar systems to work on. Participants are therefore strongly encouraged to prepare their own MPI, OpenMP and hybrid MPI+OpenMP parallel application codes for analysis. For further information, please visit the workshop web page at: http:// www.vi-hps.org/vi-hps-tw10/ For information about other VI-HPS workshops and tutorials, and to sign up to the mailing list for announcements, visit: http://www.vi-hps.org/ training/ Contact: Ventsislav Petkov (TUM) +49 89 289-17680 Brian Wylie (JSC) +49 2461 61-6589 ********************************************************************* Visit www.vi-hps.org for more information about the Virtual Institute - High Productivity Supercomputing. ************************************* ******************************** This information is also available on our web server http://www.lrz-muenchen.de/services/compute/aktuell/ali4410/ Wolfram Hesse

1 0

Upd:MPP Cluster: Returned to user operation
by LRZ aktuell Oct. 9, 2012

Oct. 9, 2012

Changes to this message: Returned to user operation The MPP cluster has been returned to user operation after a rather over-extended maintenance; apart from difficulties with integrating the new hardware there also was some fallout from the recent power interruption which required more extended repairs than expected. Note that login presently only works via lxlogin1.lrz.de. The second login node is presently not yet available. This information is also available on our web server http://www.lrz-muenchen.de/services/compute/aktuell/ali4403/ Reinhold Bader

1 0

Again problems with GPFS file system on SuperMUC
by LRZ aktuell Oct. 9, 2012

Oct. 9, 2012

We have again problems with the parallel filesystems on SuperMUC. IBM is working on this issue. Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/compute/hlrb/aktuell/ali4428/ Matthias Brehm

1 0

Again problems with GPFS file system on SuperMUC
by LRZ aktuell Oct. 9, 2012

Oct. 9, 2012

We have again problems with the parallel filesystems on SuperMUC. IBM is working on this issue. Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/compute/hlrb/aktuell/ali4427/ Matthias Brehm

1 0