HP XC System 3.x Software manuals
Computer Equipment > Software
When we buy new device such as HP XC System 3.x Software we often through away most of the documentation but the warranty.
Very often issues with HP XC System 3.x Software begin only after the warranty period ends and you may want to find how to repair it or just do some service work.
Even oftener it is hard to remember what does each function in Software HP XC System 3.x Software is responsible for and what options to choose for expected result.
Fortunately you can find all manuals for Software on our side using links below.
HP XC System 3.x Software Manual
118 pages 1.39 Mb
1 Version3 Table of Contents7 List of Figures9 List of Tables11 List of Examples13 About This DocumentIntended Audience Document Organization 14 HP XC Information15 For More InformationSupplementary Information 16 Related Information17 Related Linux Web Sites•http://www.redhat.com •http://www.linux.org/docs/index.html •http://www.linuxheadquarters.com •http://linuxvirtualserver.org •http://www.gnu.org bash emacs gdb Related MPI Web Sites•http://www.mpi-forum.org •http://www-unix.mcs.anl.gov/mpi Related Compiler Web Siteshttp://www.intel.com/software/products/compilers/index.htm Web site for Intel® compilers •http://support.intel.com/support/performancetools Web site for general Intel software development information •http://www.pgroup.com Home page for The Portland Group™, supplier of the PGI® compiler Additional Publications•Linux Administration Unleashed, by Thomas Schenk, et al , by Thomas Schenk, et al Managing NFS and NIS •MySQL, by Paul Debois •MySQL Cookbook, by Paul Debois •High Performance MySQL, by Jeremy Zawodny and Derek J. Balling (O'Reilly) 18 Typographic ConventionsHP Encourages Your Comments 19 1 Overview of the User EnvironmentSystem Architecture Table 1-1Determining the Node Platform processor vendor_id : GenuineIntel cpu family model model name : Intel(R) Xeon(TM) : AuthenticAMD : AMD Opteron(tm) vendor arch : IA-64 family : Itanium Note The /proc/cpuinfo file is dynamic 20 Node Specializationhead node of the HP XC system client nodes such as logging into the system or running jobs login role nodes with the login role in an HP XC system, depending upon cluster size and node in this manual compute role nodes in this manual Storage and I/OSAN Storage Local Storage HP XC file systems are described in detail in File System 21 File Systemimplemented as ioctls File System Layout •HP XC-specificsoftware is located in /opt/hptc •HP XC configuration data is located in /opt/hptc/etc •Clusterwide directory structure (file system) is located in /hptc_cluster 22 Determining System Configuration Information23 User Environment24 Application Development EnvironmentRun-TimeEnvironment 25 Load Sharing Facility (LSF-HPC)Standard LSF How LSF-HPCand SLURM Interact-srun JOB_STARTER scancel After a job finishes, LSF-HPCreleases all allocated resources "How and SLURM Launch and Manage a Job In summary, and in general: 26 Components, Tools, Compilers, Libraries, and Debuggers27 2 Using the SystemLogging In to the System Overview of Launching and Managing Jobs 28 IntroductionEnvironment (page 24) JobID How and SLURM Launch and Manage a Job (page 73) Getting Information About Queuesbqueues •Queue name •Queue priority •Queue status •Job slot statistics •Job state statistics To get information about queues, enter the bqueues as follows: $ bqueues Examining System Queues (page 76) Getting Information About ResourcesThe LSF $ bhosts Getting the Status of (page 75) $ lshosts Getting Information About LSF Execution Host Node (page 75) lsload 29 Performing Other Common User Tasks30 Getting System Help and Information31 3 Configuring Your Environment with Modulefiles37 4 Developing Applications45 5 Submitting Jobs53 6 Debugging ApplicationsDebugging Serial Applications Debugging Parallel Applications 54 SSH and TotalViewSetting Up TotalView module load mpimodule load totalview Using TotalView with SLURM $ srun -Nx -A $ mpirun -tv -srun application 55 Using TotalView with LSF-HPC$ bsub -nx -ext"SLURM[nodes=x]" \ -Is /usr/bin/xterm Setting TotalView Preferences $ totalview %C %R -n"%B/tvdsvr -working_directory%D -callback%L -set_pw%P -verbosity%V %F Debugging an Application $ mpicc -g -oPsimple simple.c -lm $ mpirun -tv -srun -n2 ./Psimple 56 Debugging Running Applications$ mpirun -srun -n2Psimple 57 Exiting TotalView$ squeue $ scancel --user username 59 7 Tuning Applications63 8 Using SLURM67 9 Using LSFUsing Standard LSF on an HP XC System Using LSF-HPC•Introduction to LSF-HPCin the HP XC Environment (page 68) •Determining the LSF Execution Host (page 75) •Determining Available LSF-HPCSystem Resources (page 75) •Submitting Jobs (page 77) •Getting Information About Jobs (page 80) •Translating SLURM and LSF-HPCJOBIDs (page 83) •Working Interactively Within an LSF-HPCAllocation (page 84) •LSF-HPCEquivalents of SLURM srun Options (page 86) 68 Introduction to LSF-HPCin the HP XC EnvironmentOverview of LSF-HPC •Request contiguous nodes •Execute only one task per node •Request nodes with specific features "HP XCCompute Node Resource Support "Differences Between and Standard LSF 69 Differences Between LSF-HPCand Standard LSF"Notes About Using in the HP XC Environment By LSF standards, the HP XC system is a single host. Therefore, all LSF LSF-SLURM External Scheduler (page 78) •maxswap •ndisks •r15s •r1m •r15m •ut •pg •io •tmp •swp •mem load The lshosts and lsload commands display “-”for each of these items Platform LSF HPC version number for SLURM, date stamp HOST_NAME type cpuf ncpus maxmem maxswp server RESOURCES lsfhost.loc SLINUX6 Opteron8 8 2007M (slurm) STATUS JL/U MAX NJOBS RUN SSUSP USUSP RSV lsfhost.localdomai ok 70 n15UNKNOWN server 2007M $ ssh n15 lshosts UNKNOWN UNKNOWN •LSF-HPC-enforced job-level run-timelimits are not supported -Is -Ip "Working Interactively Within an Allocation •LSF-HPCdoes not support user-accountmapping and system-accountmapping •LSF-HPCdoes not support topology-awareadvanced reservation scheduling Job Terminology The following terms are used to describe jobs submitted to LSF-HPC Batch job A job submitted to LSF or SLURM that runs without any I/O connection back to the terminal from which the job was submitted. This job may run availability and batch system scheduling policies such as files from which to read input and filenames to collect output By default, LSF jobs are batch jobs. The output is e-mailedto the user which requires that e-mailbe set up properly. SLURM batch jobs are submitted with the srun -b command. By default, the output is written to $CWD/slurm-SLURMjobID.out from the node on which the batch job was launched Use Ctrl-Cat any time to terminate the job Interactive batch job terminal from which the job was submitted. The job is also subject to resource availability and scheduling policies, so it may pause before can provide input if the job allows it with the bsub -I command Serial job constraints: mem mincpus nodes 71 HP XCCompute Node Resource Support72 $ bsub -n10 -Isrun hostname$ bsub -n10 -ext"SLURM[nodes=10]" -Isrun hostname $ bsub -n10 -ext"SLURM[nodes=10;exclude=n16]" -Isrun hostname $ bsub -n10 -ext"SLURM[constraint=dualcore]" -Isrun hostname $ bsub -n10 -ext "SLURM[nodelist=n[1-10]]"srun hostname $ bqueues -ldualcore | grep SLURM Notes on LSF-HPC $ squeue --jobs$SLURM_JOBID 73 How LSF-HPCand SLURM Launch and Manage a JobUser $ bsub -n4 -ext"SLURM[nodes=4]" -ooutput.out ./myscript 74 Job Startup and Job Control75 Preemption76 The following example shows the output from the lshosts command:lsfhost.loc SLINUX6 Itanium2 3456M UNKNOWN Of note in the lshosts output: SLINUX64 ncpus maxmem maxtmp Getting Host Load Information The LSF lsload command displays load information for LSF execution hosts status OUTPUT metrics Examining LSF-HPCSystem Queues See bsub(1) for more information on submitting jobs to specific queues For more information on the bqueues command, see bqueues(1) Getting Information About the lsf Partition $ sinfo -plsf PARTITION AVAIL TIMELIMIT NODES STATE NODELIST up infinite idle n[1-128] $ sinfo -plsf -lNe PARTITION STATE CPUS MEMORY TMP_DISK WEIGHT FEATURES n[1-128] lsf idle (null) none Refer to "Getting System Information with the sinfo Command and the sinfo(1) manpage and for further information about using the sinfo command 77 Submitting JobsChapter 5. Submitting Jobs The basic synopsis of the bsub command is: bsub [bsub-options]jobname jobname [job-options] jobname "Submitting a Parallel Job That Uses the Message Passing Interface "Enabling Remote Execution with OpenSSH Summary of the LSF bsub Command Format bsub [bsub-options] jobname [job-arguments] "Submitting a Serial Job Using bsub -n num-procs [bsub-options] jobname [job-arguments] 78 about running jobs. Refer to "Submitting a Batch Job or Job Scriptfor information about running scripts bsub -n num-procs [bsub-options]srun [srun-options] jobname [job-arguments] "Submitting a Non-MPIParallel Job bsub-n bsub -n num-procs -ext "SLURM[slurm-arguments] [bsub-options][srun [srun-options]] jobname [job-options] [srun LSF-SLURMExternal Scheduler The format for the external scheduler is: When this option is added to an LSF command line, it looks like the following: bsub -n num-procs -ext "SLURM[slurm-arguments]" [bsub-options][srun [srun-options]] jobname [job-options] 79 Job <70> is submitted to default queue <normal>. <<Waiting for dispatch<<Starting on lsfhost.localdomain>> n6 Example 9-2 shows one way to submit a parallel job to run one task per node Job <71> is submitted to default queue <normal>. <<Waiting for dispatch Job <72> is submitted to default queue <normal <<Waiting for dispatch ...>> <<Starting on lsfhost.localdomain>> n1 This example runs the job exactly the same as in , but additionally requests that node Submitting a Job from a Non-HPXC Host lsf.shared -R"type=SLINUX64 The following example shows this resource requirement string in an LSF command: $ bsub -R"type=SLINUX64" -n4 -Isrun hostname 80 Job Allocation Information for a Running Job81 Job Allocation Information for a Finished Job82 Example 9-5Using the bjobs Command (Long Output)Job <24>, User <msmith>,Project <default>,Status <RUN date and time stamp: Submitted from host <n16>, CWD <$HOME Viewing the Historical Information for a Job 9-6Using the bhist Command (Short Output) $ bhist Summary of time in seconds spent in various states: smith bin/bash 9-1 “Output Provided by the bhist Command” Table 9-1Output Provided by the bhist Command 83 Translating SLURM and LSF-HPCJOBIDs$ bsub -o%J.out -n8 sleep $ bjobs -l99 | grep slurm $ bhist -l99 | grep slurm $ scontrol show job | grep Name $ sacct -j123 CANCELLED COMPLETED •Creates the allocation in SLURM •Submits the user job to SLURM •Waits for the user job to finish •Cancels the allocation in SLURM •The first entry represents the allocation created by LSF-HPC The second entry, SLURM job step 0, represents the user job that Further entries represent 84 Working Interactively Within an LSF-HPCAllocation$ bsub -I -n4 -ext"SLURM[nodes=4]" /bin/bash <<Starting on lsfhost.localdomain <<Waiting for dispatch $ bjobs -l124 | grep slurm date and time stamp: slurm_id=150;ncpus=8;slurm_alloc=n[1-4]; n[1-4] --jobid $ srun --jobid=150hostname $ export SLURM_JOBID=150 $ srun hostname 85 $ unset SLURM_JOBIDSection (page 48) Example 9-8Launching an Interactive MPI Job Example 9-9Launching an Interactive MPI Job on All Cores in the Allocation This example assumes 2 cores per node $ mpirun -srun --jobid=150 -n8hellompi Hello! I'm rank 0 of 8 on n1 Hello! I'm rank 0 of 8 on n1 Hello! I'm rank 1 of 8 on n1 Hello! I'm rank 2 of 8 on n2 Hello! I'm rank 3 of 8 on n2 Hello! I'm rank 4 of 8 on n3 Hello! I'm rank 5 of 8 on n3 Hello! I'm rank 6 of 8 on n4 Hello! I'm rank 7 of 8 on n4 Alternatively, you can use the following: ssh X forwarding $ export SLURM_JOBID=150 $ export SLURM_NPROCS=4 $ mpirun -tvsrun additional parameters as needed mpirun -tv srun unset $ unset SLURM_NPROCS $ srun --jobid=250uptime $ bsub -n4 -ext"SLURM[nodes=4]" -o%J.out sleep 86 LSF-HPCEquivalents of SLURM srun Options91 10 Advanced Topics99 Appendix A Examples109 Glossary110 FCFSto the queue first-come See FCFS first-served global storage as local storage golden client and golden client golden image are not appropriate for replication golden master solely by the use of this collection of directories and files for the head node, but HP XC system has only one head node at any one time host name characters, with the first character being a letter I/O node storage within the HP XC system iLO hardware models that enables remote management of any node within a system image server client Integrated Lights See iLO Out interconnect applications communications within applications Internet address similar to the following: IPMI 111 Linux VirtualSee LVS Server load file single command Load Sharing See LSF-HPCwith SLURM Facility local storage Storage that is available or accessible from one node in the HP XC system LSF execution host from the LSF execution host and are executed on one or more compute nodes LSF master host becomes the master host LSF-HPCwith processing with SLURM on the LSF execution host See Also LSF execution host LVS login requests and directs them to a node with a login role Management See MP Processor master host See LSF master host module modulefiles See Also modulefile Modulefiles enable various functions to start and operate properly on HP Integrity servers MPI by a broadly based committee of vendors, implementors, and users MySQL and track system configuration information NAT two machines to different external IP addresses Network Address See NAT Translation 112 NetworkSee NIS Information Services NIS across a set of networked hosts NIS client maintain copies of the NIS maps locally for their domain NIS master server the master maps can be modified. Each domain can have only one master server NIS slave server distributed throughout the network parallel application library PXE remotely. PXE booting is configured at the BIOS level resource Nodes with this role manage the allocation of resources to user applications management role role A set of services that are assigned to a node Root Administration Switch is connected to the Root Administration Switch root node RPM Red Hat Package Manager notably to install and remove software packages package management utility serial application calls (for example, MPI, PVM, GM, or Portals) infrastructure libraries SLURM backup controller becomes the SLURM master controller SLURM master The node on which the slurmctld daemon runs 113 SMPpower) available per unit of space ssh network standard LSF single-processjobs. Standard LSF is not integrated with SLURM symmetric See SMP multiprocessing 115 Index
2 pages 37.43 Kb
1 Sample UNIX installation directoriesDaemon error log filesLSF base system daemon log files LSF batch system daemon log files Configuration files Cluster configuration parameters (lsf.conf)Variable UNIX Default Version Administration and accounting commands Daemons 2 User commandsViewing information about your cluster Monitoring jobs and tasks Submitting and controlling jobs bsub command
Also you can find more HP manuals or manuals for other Computer Equipment.