- aktuell - lists.lrz.de

Netzwartung am 5.2.2013 / 12.2.2013
by LRZ aktuell Feb. 4, 2013

Feb. 4, 2013

Am 5.2. sind vom LRZ keine Wartungarbeiten vorgesehen. Der DFN-Verein hat fuer den X-WiN-Anschluss des Schneefernerhauses auf der Zugspitze eine Router-Wartung zwischen 2:00 Uhr und 5:30 Uhr angekuendigt. Vorankuendigung: Naechste Woche, am 12.2. ist die Ersetzung des Routers csr1-kic in Grosshadern geplant. Dudurch wird das Netz im Campus Grosshadern/Martinsried zwischen 7:30 Uhr und 8:00 Uhr ausfallen. Es wird der erste Einsatz eines Routers der Plattform Cisco Nexus 7000 im MWN sein. Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/netz/aktuell/ali4516/ Wolfgang Beyer

1 0

Upd:Linux Cluster: Maintenance and Configuration Changes Jan 28 - Feb 1
by LRZ aktuell Feb. 1, 2013

Feb. 1, 2013

Changes to this message: Partially returned to user operation Update (Feb 1 17:15): Except for the MPP segment, the cluster systems have been returned to user operation. Possible configuration problems on the MPP cluster still need investigation, please bear with us. ----------------------------------------------------------------------- Maintenance windows Due to an extended hard- and software maintenance, the MPP, MAC and 8-way Myrinet clusters will be unavailable for user operation between Jan 28, 8:00 am and Feb 1. This includes the login nodes lxlogin1-2. ( Note: Until Jan 31, 8:00am, the lxlogin3.lrz.de system will remain available for submission of serial jobs.) All other cluster segments (serial cluster including all hosted systems, sgi ICE and UV systems, 32 way Myrinet systems) will be unavailable between Jan 31, 8:00 am and Feb 1. Configuration changes * The default MPI environment on the MPP cluster will be changed from Parastation MPI to Intel MPI. However, the mpi.parastation module will remain available for legacy use until the end of 2013. On the sgi ICE and UV systems, the sgi MPI (mpi.mpt) will remain default. * The 8-way Myrinet Cluster will be retired from parallel processing, and the nodes will be added to the serial processing pool. This implies that the partition "myri_std" in the SLURM cell "myri" will become unavailable. * For the serial queues, SLURM fair share scheduling will be introduced. For the parallel queues, a combination of fair share scheduling and favoring large jobs will be activated. This is to prevent a single user from monopolizing cluster segments for long times if there are many jobs in the queue. * New storage systems will be introduced for the WORK (==PROJECT) and SCRATCH file systems. Please note that LRZ will only migrate WORK data to the new file system; data in SCRATCH will not be migrated. However the old SCRATCH file system will remain available as a separate mount in read-only mode on the login nodes until the end of March, 2013. Migration of data can then be done via commands like cd $SCRATCH_LEGACY cp -a my_scratch_subdirectory $SCRATCH The environment variable $SCRATCH_LEGACY will remain defined and point to the legacy scratch area until end of March, 2013. Further configuration changes may occur and will be added to the above list if they are applied. This information is also available on our web server http://www.lrz-muenchen.de/services/compute/aktuell/ali4501/ Reinhold Bader

1 0

Upd:SuperMUC: scheduled maintenance Jan 28 31
by LRZ aktuell Feb. 1, 2013

Feb. 1, 2013

Changes to this message: Extended maintenance continues Update 3 (Feb 1 14:50): IBM is still investigating the cause of the problems observed in internal testing. It appears that there exists an incompatibility between one or more updated system software components. Therefore, no user operation can be expected for the weekend. We will update you on the status on Monday, Feb 4. ----------------------------------------------------------------------- Update (Jan 30 afternoon): In order to check additional hardware and firmware modifications that were done to improve the stability of the system (see also http://www.lrz.de/services/compute/supermuc/aktuell/ ali4512/, as well as dedicated operation for a project that requires short-term use of the complete machine, user operation of the system is delayed to the afternoon of January 31. ----------------------------------------------------------------------- Dear users of the SuperMUC petascale system at LRZ, Due to a combined hard- and software maintenance, SuperMUC will be unavailable for user operation between January 28, 8:00 and the afternoon of January 31. Jobs still running at the beginning of the maintenance will be cancelled. The following changes are introduced by this maintenance: * Access to the PRACE network will become available to PRACE users, * LoadLeveler mail notification will be activated, * Licence servers will be accessible from the compute nodes, * The Infiniband software stack will be updated to a newer release, * Some presently missing packages will become available on the compute nodes (e.g., you can remove the workaround for a missing libnuma.so) This information is also available on our web server http://www.lrz-muenchen.de/services/compute/supermuc/aktuell/ali4502/ Reinhold Bader

1 0

Upd:Linux Cluster: Maintenance and Configuration Changes Jan 28 - Feb 1
by LRZ aktuell Feb. 1, 2013

Feb. 1, 2013

Changes to this message: Updated changes Maintenance windows Due to an extended hard- and software maintenance, the MPP, MAC and 8-way Myrinet clusters will be unavailable for user operation between Jan 28, 8:00 am and Feb 1. This includes the login nodes lxlogin1-2. ( Note: Until Jan 31, 8:00am, the lxlogin3.lrz.de system will remain available for submission of serial jobs.) All other cluster segments (serial cluster including all hosted systems, sgi ICE and UV systems, 32 way Myrinet systems) will be unavailable between Jan 31, 8:00 am and Feb 1. Configuration changes * The default MPI environment on the MPP cluster will be changed from Parastation MPI to Intel MPI. However, the mpi.parastation module will remain available for legacy use until the end of 2013. On the sgi ICE and UV systems, the sgi MPI (mpi.mpt) will remain default. * The 8-way Myrinet Cluster will be retired from parallel processing, and the nodes will be added to the serial processing pool. This implies that the partition "myri_std" in the SLURM cell "myri" will become unavailable. * For the serial queues, SLURM fair share scheduling will be introduced. For the parallel queues, a combination of fair share scheduling and favoring large jobs will be activated. This is to prevent a single user from monopolizing cluster segments for long times if there are many jobs in the queue. * New storage systems will be introduced for the WORK (==PROJECT) and SCRATCH file systems. Please note that LRZ will only migrate WORK data to the new file system; data in SCRATCH will not be migrated. However the old SCRATCH file system will remain available as a separate mount in read-only mode on the login nodes until the end of March, 2013. Migration of data can then be done via commands like cd $SCRATCH_LEGACY cp -a my_scratch_subdirectory $SCRATCH The environment variable $SCRATCH_LEGACY will remain defined and point to the legacy scratch area until end of March, 2013. Further configuration changes may occur and will be added to the above list if they are applied. This information is also available on our web server http://www.lrz-muenchen.de/services/compute/aktuell/ali4501/ Reinhold Bader

1 0

Upd:SuperMUC: scheduled maintenance Jan 28 31
by LRZ aktuell Jan. 31, 2013

Jan. 31, 2013

Changes to this message: Prolongation of maintenance Update 2 (Jan 31 17:50): Internal testing unfortunately has revealed a number of problems that prevent us from returning the SuperMUC system to user operation. We intend to provide daily updates on the status of the system. ----------------------------------------------------------------------- Update (Jan 30 afternoon): In order to check additional hardware and firmware modifications that were done to improve the stability of the system (see also http://www.lrz.de/services/compute/supermuc/aktuell/ ali4512/, as well as dedicated operation for a project that requires short-term use of the complete machine, user operation of the system is delayed to the afternoon of January 31. ----------------------------------------------------------------------- Dear users of the SuperMUC petascale system at LRZ, Due to a combined hard- and software maintenance, SuperMUC will be unavailable for user operation between January 28, 8:00 and the afternoon of January 31. Jobs still running at the beginning of the maintenance will be cancelled. The following changes are introduced by this maintenance: * Access to the PRACE network will become available to PRACE users, * LoadLeveler mail notification will be activated, * Licence servers will be accessible from the compute nodes, * The Infiniband software stack will be updated to a newer release, * Some presently missing packages will become available on the compute nodes (e.g., you can remove the workaround for a missing libnuma.so) This information is also available on our web server http://www.lrz-muenchen.de/services/compute/supermuc/aktuell/ali4502/ Reinhold Bader

1 0

Gauss Call #9 for Large Scale Projects
by LRZ aktuell Jan. 31, 2013

Jan. 31, 2013

Gauss Call #9 for Large-Scale Projects see also: Current call as PDF (http://www.gauss-centre.eu/datapool/page /71/GaussCall9.pdf) Large-scale projects and highly scalable parallel applications are characterised by large computing time requirements, not only for short time frames, but often for longer time periods. Projects are classified as Large-Scale , if they require, in general, more than 2% of the potentially available core cycles for GCS in one year on a member centre s high-end system, i.e. * more than 36 Mill. core-hours per year at JSC (IBM Blue Gene/Q), * more than 20 Mill. core-hours per year at LRZ (IBM iDataPlex SuperMUC), * more than 16 Mill. core-hours per year at HLRS (Cray HERMIT Step 1). For these large-scale projects a competitive review and resource allocation process is established by the GCS. Requests for resources below these limits will be accepted by the individual member centres. Requests above these limits will be handed over to GCS, handled according to the joint procedures, and will be reviewed in a national context. A "Call for Large-Scale Projects" is published by the Gauss Centre twice a year. Dates for closure of calls are usually at the end of winter and at the end of summer of each year. Eligible are applications from German universities and publicly funded German research institutions, e.g., Max-Planck Society, and Helmholtz Association. The deadline for the current call is: 28 February 2013, 17:00. Answering the Call Leading, ground-breaking projects should deal with complex, demanding, innovative simulations that would not be possible without the GCS infrastructure, and which can benefit from the exceptional resources provided by GCS. Application for a large-scale project should be performed by filling in the appropriate electronic application form that can be accessed from the GCS web page "Computing Time" (http://www.gauss-centre.eu/computing-time) Note that the regular application forms of the GCS member centres can be reached from there, and users are required to choose a centre that provides the architecture and software most appropriate for their programs. The proposals for large-scale projects will be first reviewed with respect to their technical feasibility. Afterwards, they will be peer-reviewed for a comparative scientific evaluation. On the basis of this evaluation by a GCS committee the projects will be approved for a period of one year and given their allocations. A project with a running large-scale grant on one of the centres or targeting multiple GCS platforms must clearly indicate and justify this as part of the application. Criteria for decision Applications for compute resources are evaluated only according to their scientific excellence. * The proposed scientific tasks must be scientifically challenging, and their treatment must be of substantial interest. * Clear scientific goals and verifiable milestones on the way to reach these goals must be specified. * The implementation of the project must be technically feasible on the available computing systems, and must be in reasonable proportion to the performance characteristics of these systems. * The Principal Investigator must have a proven scientific record, and she/he must be able to successfully accomplish the proposed tasks. In particular, applicants must possess the necessary specialized know-how for the effective use of high-end computing systems. This has to be proven in the application for compute resources, e.g. by presenting work done on smaller computing system, scaling studies etc. * The specific features of the high-end computers should be optimally exploited by the program implementations. This will be checked regularly during the course of the project. Application for Large-Scale Projects for SuperMUC Project applications for SuperMUC may be submitted by any scientist qualified in his or her respective field of research. Computing resources are allocated on the basis of independent referees' reports. Apart from the scientific relevance of the project, an important criterion for the allocation of computing resources is that the project can make efficient use of the computer and use a large number of processors in parallel for the simulations. Please, first prepare a detailed project description that you will have to upload to the electronic questionnaire as a pdf file. Follow the guidelines (https://www.gauss-centre.eu/computing-time/supermuc/ guidelines) about form, content and size of the project description. Take particular care that * the estimate of the applied computing time is verifiable by the reviewer, * a plot is supplied that shows the scalability of the program under realistic production conditions, and that * you provide information on computing projects and applications for computing time you have with other supercomputing centres. If you wish to add supplemental material to the project description (unpublished manuscripts etc.), please collect that in a second pdf-file. Also this file you will have to upload to the electronic questionnaire. Please fill in the electronic questionnnaire completely and attach the project description and - if necessary - the supplemental material as pdf files. * Electronic questionnaire (https://pound.zam.kfa-juelich.de/ lrz_projekt/) (open 30 January- 28 February 2013, 17:00) * Show, change, print electronic questionnaire (https:// pound.zam.kfa-juelich.de/lrz_prodis/) Please make sure that your data are complete and check them. Finally, please print the application form and send a signed paper copy to : GCS Large-Scale Projects c/o NIC Coordination office Juelich Supercomputing Centre Forschungszentrum Juelich 52425 Juelich Germany This information is also available on our web server http://www.lrz-muenchen.de/services/compute/supermuc/aktuell/ali4515/ Matthias Brehm

1 0

Gauss Call #9 for Large-Scale Projects
by LRZ aktuell Jan. 30, 2013

Jan. 30, 2013

Gauss Call #9 for Large-Scale Projects see also: Current call as PDF (http://www.gauss-centre.eu/datapool/page /71/GaussCall9.pdf) Large-scale projects and highly scalable parallel applications are characterised by large computing time requirements, not only for short time frames, but often for longer time periods. Projects are classified as Large-Scale , if they require, in general, more than 2% of the potentially available core cycles for GCS in one year on a member centre s high-end system, i.e. * more than 36 Mill. core-hours per year at JSC (IBM Blue Gene/Q), * more than 20 Mill. core-hours per year at LRZ (IBM iDataPlex SuperMUC), * more than 16 Mill. core-hours per year at HLRS (Cray HERMIT Step 1). For these large-scale projects a competitive review and resource allocation process is established by the GCS. Requests for resources below these limits will be accepted by the individual member centres. Requests above these limits will be handed over to GCS, handled according to the joint procedures, and will be reviewed in a national context. A "Call for Large-Scale Projects" is published by the Gauss Centre twice a year. Dates for closure of calls are usually at the end of winter and at the end of summer of each year. Eligible are applications from German universities and publicly funded German research institutions, e.g., Max-Planck Society, and Helmholtz Association. The deadline for the current call is: 28 February 2013, 17:00. Answering the Call Leading, ground-breaking projects should deal with complex, demanding, innovative simulations that would not be possible without the GCS infrastructure, and which can benefit from the exceptional resources provided by GCS. Application for a large-scale project should be performed by filling in the appropriate electronic application form that can be accessed from the GCS web page "Computing Time" (http://www.gauss-centre.eu/computing-time) Note that the regular application forms of the GCS member centres can be reached from there, and users are required to choose a centre that provides the architecture and software most appropriate for their programs. The proposals for large-scale projects will be first reviewed with respect to their technical feasibility. Afterwards, they will be peer-reviewed for a comparative scientific evaluation. On the basis of this evaluation by a GCS committee the projects will be approved for a period of one year and given their allocations. A project with a running large-scale grant on one of the centres or targeting multiple GCS platforms must clearly indicate and justify this as part of the application. Criteria for decision Applications for compute resources are evaluated only according to their scientific excellence. * The proposed scientific tasks must be scientifically challenging, and their treatment must be of substantial interest. * Clear scientific goals and verifiable milestones on the way to reach these goals must be specified. * The implementation of the project must be technically feasible on the available computing systems, and must be in reasonable proportion to the performance characteristics of these systems. * The Principal Investigator must have a proven scientific record, and she/he must be able to successfully accomplish the proposed tasks. In particular, applicants must possess the necessary specialized know-how for the effective use of high-end computing systems. This has to be proven in the application for compute resources, e.g. by presenting work done on smaller computing system, scaling studies etc. * The specific features of the high-end computers should be optimally exploited by the program implementations. This will be checked regularly during the course of the project. Application for Large-Scale Projects for SuperMUC Project applications for SuperMUC may be submitted by any scientist qualified in his or her respective field of research. Computing resources are allocated on the basis of independent referees' reports. Apart from the scientific relevance of the project, an important criterion for the allocation of computing resources is that the project can make efficient use of the computer and use a large number of processors in parallel for the simulations. Please, first prepare a detailed project description that you will have to upload to the electronic questionnaire as a pdf file. Follow the guidelines (https://www.gauss-centre.eu/computing-time/supermuc/ guidelines) about form, content and size of the project description. Take particular care that * the estimate of the applied computing time is verifiable by the reviewer, * a plot is supplied that shows the scalability of the program under realistic production conditions, and that * you provide information on computing projects and applications for computing time you have with other supercomputing centres. If you wish to add supplemental material to the project description (unpublished manuscripts etc.), please collect that in a second pdf-file. Also this file you will have to upload to the electronic questionnaire. Please fill in the electronic questionnnaire completely and attach the project description and - if necessary - the supplemental material as pdf files. * Electronic questionnaire (https://pound.zam.kfa-juelich.de/ lrz_projekt/) (open 30 January- 28 February 2013, 17:00) * Show, change, print electronic questionnaire (https:// pound.zam.kfa-juelich.de/lrz_prodis/) Please make sure that your data are complete and check them. Finally, please print the application form and send a signed paper copy to : GCS Large-Scale Projects c/o NIC Coordination office Juelich Supercomputing Centre Forschungszentrum Juelich 52425 Juelich Germany Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/compute/hlrb/aktuell/ali4514/ Matthias Brehm

1 0

Upd:SuperMUC: scheduled maintenance Jan 28 31
by LRZ aktuell Jan. 30, 2013

Jan. 30, 2013

Changes to this message: Additional time needed for testing Update: In order to check additional hardware and firmware modifications that were done to improve the stability of the system (see also http://www.lrz.de/services/compute/supermuc/aktuell/ali4512/, as well as dedicated operation for a project that requires short-term use of the complete machine, user operation of the system is delayed to the afternoon of January 31. ----------------------------------------------------------------------- Dear users of the SuperMUC petascale system at LRZ, Due to a combined hard- and software maintenance, SuperMUC will be unavailable for user operation between January 28, 8:00 and the afternoon of January 31. Jobs still running at the beginning of the maintenance will be cancelled. The following changes are introduced by this maintenance: * Access to the PRACE network will become available to PRACE users, * LoadLeveler mail notification will be activated, * Licence servers will be accessible from the compute nodes, * The Infiniband software stack will be updated to a newer release, * Some presently missing packages will become available on the compute nodes (e.g., you can remove the workaround for a missing libnuma.so) This information is also available on our web server http://www.lrz-muenchen.de/services/compute/supermuc/aktuell/ali4502/ Reinhold Bader

1 0

TSM-Kernel-Update am 04.02.2013 von 12 Uhr bis ca. 14 Uhr
by LRZ aktuell Jan. 30, 2013

Jan. 30, 2013

Wegen eines dringenden Kernel-Updates sind die TSM-Server S39 - S56, S60, S62, S65, S67-S68 am Montag, den 04.02.2013 in der Zeit von 12:00 bis ca. 14:00 Uhr jeweils fuer ca. 30 Minuten nicht verfuegbar. Zudem wird es in dieser Zeit eine 30 minuetige Unterbrechung bei den Tape-Mounts des SuperMUC Archivs (T68I1 - T68I4, T69I1 - T69I4) geben. Wir bitten um Ihr Verstaendnis. Due to a high-urgency kernel update the TSM servers S39 - S56, S60, S62, S65, S67-S68 will be unavailable for approximately half an hour on Monday, Feb 4, 2013, between 12:00 and ca 14:00. During this time, tape mounts will not be executed for half an hour also on the SuperMUC archive (servers T68I1 - T68I4, T69I1 - T69I4). With apologies for the disruption. Ihr ABS-Team Diese Information finden Sie im WWW unter http://www.lrz-muenchen.de/services/datenhaltung/adsm/aktuell/ali4513/ Monika Hufnagl

1 0

SuperMUC: On the way to resolving the stability issues
by LRZ aktuell Jan. 30, 2013

Jan. 30, 2013

Dear users of the SuperMUC Petaflop system, As known to many of you, we have encountered significant stability issues on the system throughout the last months, not all of which were resolved up to January 2013. These problems particularly impact large jobs, as well as jobs that highly stress the I/O subsystems, especially GPFS. Investigation of the causes for these instabilities has progressed to a point that allows us to provide information on IBM s and LRZ s plans for significant improvement of operational stability. Issue 1: Occasional switch outages caused by a race condition in the management part of the switch logic. This issue will be fixed during the Jan 28-31 maintenance phase. Issue 2: A relatively high number of Infiniband cables has shown degradation or failure. While the exact cause is still under investigation, the prime suspect is a too high temperature in some parts of the air-cooled switch infrastructure that will be resolved by increasing the fan speeds, to be implemented during the Jan 28-31 maintenance phase. Also, the defective cables will be replaced. Issue 3: The hard disks underlying the GPFS are not designed for uninterrupted high-bandwidth I/O and may degrade in performance over time if overstressed. Given the I/O profiles we observe, this is not an issue if such disks are automatically retired from usage once the problem arises, and then exchanged. However, this automatism inside the storage does not presently work for some scenarios. Unfortunately, even a small number of slow disks will considerably slow down the complete file system. At present, manual disk reviews are done in short intervals to alleviate this problem; we expect that a full solution to the problem will become available within a few weeks. Given the above, considerable stability improvements should already be observable once the system is returned to regular operation after the above-mentioned maintenance. Please provide feedback (via our Service Desk) as soon as possible if you feel this not to be the case, and also if you already have an incident ticket open at our site. This information is also available on our web server http://www.lrz-muenchen.de/services/compute/supermuc/aktuell/ali4512/ Reinhold Bader

1 0