HP High Performance LC Cluster series Manuale di Installazione

ClusterPack Index of Tutorial Sections Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Administrators Gu

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyrepeat the inst

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary be added to the cluster using the IP address and hostname

Copyright 1994-2004 hewlett-packard company

Job Management Tasks ClusterPack Job Management Tasks Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary 2

zGo to the following URL in the web browser: % /opt/netscape/netscape http://<management_server>:8080/Platform/login/Login.jsp zEnter your Unix

Using the Clusterware Pro V5.1 Web Interface: From the jobs tab: zSelect Job->Submit. zEnter job data. zClick Submit. Data files required for th

Using the Clusterware Pro V5.1 CLI: % bsub -q <queue_name> <cmd> Use bqueues to list available Queues. % bqueues References: z3.7.8 How do

Using the Clusterware Pro V5.1 Web Interface: Set a priority at submission by: zFrom the Jobs Tab, select Job->Submit. zUsing the Queue pull down

References: z3.7.8 How do I access the Clusterware Pro V5.1 Web Interface? z3.7.9 How do I access the Clusterware Pro V5.1 Command Line Interface? B

Using the Clusterware Pro V5.1 Web Interface: From the Jobs tab: zSelect Job->Submit. zClick Advanced. zSelect "Send email notification whe

Using the Clusterware Pro V5.1 Web Interface: From the Jobs tab: zSelect Tools->Find. zSelect User from the Field list. zType the user name in th

ClusterPack General Overview ClusterPack ClusterPack General Overview Index | Administrators Guide | Users Guide | Tool Overview | Related Documents

z3.7.8 How do I access the Clusterware Pro V5.1 Web Interface? z3.7.9 How do I access the Clusterware Pro V5.1 Command Line Interface? Back to Top 2

z3.7.8 How do I access the Clusterware Pro V5.1 Web Interface? z3.7.9 How do I access the Clusterware Pro V5.1 Command Line Interface? Back to Top 2

Using the Clusterware Pro V5.1 CLI: % bresume <job_ID> References: z3.7.8 How do I access the Clusterware Pro V5.1 Web Interface? z3.7.9 How do

From the Jobs tab: zSelect Tools->Find. zSelect the Advanced tab. zSelect User from the Field list in the Define Criteria section. zType the use

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary 2.1.21 Suspend a submitted MPI job Using the Clusterware

Copyright 1994-2004 hewlett-packard company

File Transfer Tasks ClusterPack File Transfer Tasks Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary 2.2

Back to Top 2.2.2 Transfer a file from intranet to all Compute Nodes in the cluster If the cluster is a Guarded Cluster, this operation is done in tw

< Copies the remote file to the local file after the job completes. Overwrites the local file if it exists. % bsub -f <local_file> < <

zFTP the file from the Head node to the external target. References: zGuarded Cluster Back to Top 2.2.5 Transfer a file from a Compute Node to anoth

of Gigabit Ethernet or Infiniband. The common components of a cluster are: zHead Node - provides user access to the cluster. In smaller clusters, the

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyFor more detail

Miscellaneous Tasks ClusterPack Miscellaneous Tasks Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary 2.3

Using the Clusterware Pro V5.1 Web Interface: From the Jobs tab: zSelect Jobs->Submit. zEnter job information. zClick Advanced. zOn the Advanced

2.3.3 Check Queue status Using the Clusterware Pro V5.1 Web Interface: From the Jobs tab: zReview the Queues table. Use the Previous and Next buttons

and should not be used while AppRS jobs are running. % apprs_clean all For jobs submitted to non-AppRS queues, the user's job submission script

#APPRS TARGETUTIL 1.0 #APPRS TARGETTIME 10 #APPRS REDUNDANCY 4 # Your job goes here: if [ "$APPRS_RESTART" = "Y" ]; then # job as

2.3.6 Restart application from a checkpoint if a Compute Node crashes If a Compute Node crashes, jobs submitted to an AppRS queue will automatically

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company% bhist <job

Cluster Management Utility Zone Overview ClusterPack Cluster Management Utility Zone Overview Index | Administrators Guide | Users Guide | Tool Over

3.1.2 What are the Easy Install Tools? The ClusterPack suite includes a set of utilities for setting up a cluster of Itanium 2 nodes. The tools manan

latency and higher bandwidth. A cluster LAN is also configured to separate the system management traffic from application message passing and file ser

zsysimage_create zsysimage_register zsysimage_distribute These scripts use ClusterPack's knowledge of the cluster configuration to simplify th

new command will not begin until the previous one is finished, i.e. these do not run in parallel. Sending a SIGINT (usually a ^C) will cause the curre

zsingle local to single local % clcp src dst zsingle local to multiple local % clcp src dst.%h zsingle local to multiple remote % clcp src dst:%h or c

Make necessary changes. % clcp checklist.%c %h:/etc/checklist which maps to: % rcp host0:/etc/checklist checklist.0 % rcp host1:/etc/checklist checkli

cluptime is used as follows: % cluptime [ [-C] cluster-group] For more details on the usage of cluptime, invoke the command: % man cluptime Back to To

3.1.10 clinfo - Shows nodes and cluster information. The clinfo command lists which hosts make up a cluster. By default, with no arguments, the curre

core tools of ClusterPack, including PCC ClusterWare Pro™ and the HP Systems Insight Manager. Node groups are collections of nodes that are subsets of

% clgroup -l group1 For more details on the usage of clgroup, invoke the command: % man clgroup Back to Top 3.1.12 clbroadcast - Telnet and MP based b

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company The clpower ut

Service ControlManager (SCM) Overview ClusterPack Service ControlManager (SCM) Overview Index | Administrators Guide | Users Guide | Tool Overview |

Version 2.0. The ClusterPack has a server component that runs on a Management Server, and client agents that run on the managed Integrity compute serv

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company3.2.2 How to in

System Inventory Manager Overview ClusterPack System Inventory Manager Overview Index | Administrators Guide | Users Guide | Tool Overview | Related

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyzThe filtering

Application ReStart (AppRS) Overview ClusterPack Application ReStart (AppRS) Overview Index | Administrators Guide | Users Guide | Tool Overview | R

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyTo use AppRS, u

Cluster Management Utility (CMU) Overview ClusterPack Cluster Management Utility (CMU) Overview Index | Administrators Guide | Users Guide | Tool Ov

3.5.3 Nodes monitoring zCluster monitoring Enhanced monitoring capabilities for up to 1024 nodes in a single window (with vertical scrollbars). zMoni

window enabled. CMU will display the last monitored logical group. Note: When starting the CMU window for the first time, the monitoring action is pe

{Terminal Server Configuration {PDU Configuration {Network Topology Adaptation {Node Management {Event Handling Configuration Back to Top 3.5.7 M

A contextual menu window appears with a right click on a node displayed in the central frame of the main monitoring CMU window. The following menu op

The Data Dictionary contains definitions for common terms that are used through the tutorial. Back to Top 1.1.3 What is the best order to review the

Many management actions such as boot, reboot, halt, or monitoring will be applied to all of the selected nodes. zHalt This sub-menu allows a system ad

before booting a node. zReboot This sub-menu allows a system administrator to issue the reboot command on all of the selected nodes. The reboot comma

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyTo improve the

NAT/IPFilter Overview ClusterPack NAT/IPFilter Overview Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyIP Aliasing or

Platform Computing Clusterware Pro V5.1 Overview ClusterPack Platform Computing Clusterware Pro V5.1 Overview Index | Administrators Guide | Users G

zOrganizations experience increased productivity from transparent single system, cluster-as-server access to compute resources. zThe Platform Computi

Setup and Configuration of a DEMO license The use of a DEMO license file (license.dat) for Clusterware Pro, as part of the ClusterPack V2.4 Clusterwar

The /etc/exports file on the Management Server, and the /etc/fstab file on each Compute Node is updatedautomatically by ClusterPack. Back to Top 3.7.

To START services on the Management Server Issue the following command on the Management Server as the super user (i.e. root): % /share/platform/clust

a link to the printable version at the bottom of the page. References: zPrintable Version Back to Top 1.1.4 Operating System and Operating Environme

% /share/platform/clusterware/lbin/cwagent stop References: z3.1.5 clsh - Runs commands on one, some, or all nodes in the cluster. Back to Top 3.7.6

zThe username and password are the same as for any normal user account on the Management Server. References: z3.7.6 How do I start and stop the Clust

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company% badmin reconf

Management Processor (MP) Card Interface Overview ClusterPack Management Processor (MP) Card Interface Overview Index | Administrators Guide | Users

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company{Step 11 Run mp

HP Systems Insight Manager (HPSIM) Overview ClusterPack HP Systems Insight Manager (HPSIM) Overview Index | Administrators Guide | Users Guide | Too

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary conditions automatically through automated event handling.

Copyright 1994-2004 hewlett-packard company

Related Documents ClusterPack Related Documents Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary 4.1.1 H

http://docs.hp.com/en/5990-8540/index.html ServiceControl Manager Troubleshooting Guide http://docs.hp.com/en/5187-4198/index.html Back to Top 4.1.3 H

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyBack to Top

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyClusterPack D

Cluster LAN/Switch A Cluster LAN/Switch is usually an Ethernet network used to monitor and control all the major system components. May also handle

Interconnect Switch An Interconnect Switch provides high speed connectivity between Compute Nodes. Used for message passing and remote memory access

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company Back to Top

Comprehensive Install Instructions ClusterPack Comprehensive Install Instructions Index | Administrators Guide | Users Guide | Tool Overview | Relat

Processor. {Verify the Management Server and the initial Compute Node. zConfigure the remaining Compute Nodes with a Golden Image. {Create a Golden

Back to Top Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyDic

Note: You will not be able to complete the following steps if you have not collected all of this information. Details At various points during the c

zHP-UX 11i Ignite-UX zHP-UX 11i V2.0 TCOE ClusterPack depends on certain open source software which is normally installed as a part of the operatine

z/var - 4GB z/share - 500MB (Clusterware edition only) Details Allocate space for these file systems when you do a fresh install of HP-UX on the Ma

Step 4 Obtain a License File Background For ClusterPack Base Edition, please refer to the Base Edition License certificate for instructions on redee

Background This document does not cover hardware details. It is necessary, however, to make certain hardware preparations in order to run the softwar

% /opt/clusterpack/bin/manager_config Back to Top Step 7 Configure the ProCurve Switch Background The ProCurve Switch is used for the management n

% > lcd /tmp % > get cpack.lic % > bye Back to Top Step 9 Install ClusterPack on the Management Server Background The ClusterPack softwa

Note: You cannot be in the /mnt/dvdrom directory when you try to mount. You will get a file busy error. When you are finished, on the local machine: 6

zUsing the ClusterPack DVD, mount and register the DVD as a software depot. zInstall the ClusterPack Manager software (CPACK-MGR) on the Management S

Cluster Management Software components after reboots. zConfigure Cluster Management Software tools. The Management Server components of HP System Man

ClusterPack Install QuickStart ClusterPack ClusterPack Install QuickStart Index | Administrators Guide | Users Guide | Tool Overview | Related Docum

manager_config Invocation manager_config is an interactive tool that configures the Management Server based on some simple quer(most of the queries ha

When you telnet to an MP, you will initially access the console of the associated server. Other options suas remote console access, power management,

console port on the MP card of each Compute Node. When you are ready to run mp_register, use this command: % /opt/clusterpack/bin/mp_register Back to

When booting a node, clbootnodes will answer the first boot questions rather than having to answer themmanually. The questions are answered using the

Background This tool is the driver that installs and configures appropriate components on every Compute Node. zRegisters Compute Nodes with HP Syst

Execute the following command. % /opt/clusterpack/bin/compute_config Back to Top Step 14 Set up HyperFabric (optional) Background The utility clne

ClusterPack can configures IP over InfiniBand (IPoIB) if the appropriate InfiniBand drivers are installed the systems. Overview If the InfiniBand IP

The finalize_config tool can be run at any time to validate the cluster configuration and to determine if there are any errors in the ClusterPack soft

LSF jobs while the archive is being made: % badmin hclose <hostname> zIn addition, you should either wait until all running jobs complete, or su

Overview zRegister the image. zDistribute the image to selected nodes. Details To distribute a Golden Image to a set of Compute Nodes, you need to

Step Q1 Fill Out the ClusterPack Installation Worksheet Print out this form and fill out all information for each node in your cluster. Installatio

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company"-a"

Installation and Configuration of Optional Components ClusterPack Installation and Configuration of Optional Components Index | Administrators Guide

Nodes in a private IP sub-net (10.x.y.z range, 192.168.p.q range), which also alleviates the need for numerous public IP addresses. IP Aliasing or Ne

HP-UX IPFilter Validation HP-UX IPFilter is installed with the default HP-UX 11i V2 TCOE bundle. To validate its installation, runthe following comman

% man 8 ipf zList the input output filter rules % ipfstat -hio Setup the NAT rules In this section, we will walk through the steps of setting up HP-UX

map lan0 192.168.0.4/32 -> 15.99.84.23/32 portmap tcp/udp 40000:60000 map lan0 192.168.0.4/32 -> 15.99.84.23/32 EOF More examples of NAT and oth

If there is no packet loss, then NAT is enabled. zDISPLAY Server Interaction Test 1. On the Compute Node, set the DISPLAY variable to a display server

The default use model of an ClusterPack cluster is that end users will submit jobs remotely through the ClusterWare GUI or by using the ClusterWare C

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyALL:ALL@<Man

Software Upgrades and Reinstalls ClusterPack Software Upgrades and Reinstalls Index | Administrators Guide | Users Guide | Tool Overview | Related D

The Compute Nodes must have Management Processor (MP) cards. ClusterPack depends on certain open source software which is normally installed as a par

nature can only be accomplished by a complete re-configuration of the cluster (See Initial Installation and Setup). The reinstallation path is only me

1.4.4 Upgrading from Base Edition to Clusterware Edition Upgrading from Base Edition to Clusterware Edition is done using the "forced reinstall&

This tool is the main installation and configuration driver. Invoke this tool with "force install" option -F: % /opt/clusterpack/bin/manager

1.4.5 Upgrading from V2.2 to V2.4 ClusterPack V2.4 supports an upgrade path from ClusterPack V2.2. Customers that currently deploy ClusterPack V2.2

% /opt/clusterpack/bin/compute_config -u zVerify that everything is working as expected. % /opt/clusterpack/bin/finalize_config Back to Top 1.4.6 Upg

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyzVerify that ev

Golden Image Tasks ClusterPack Golden Image Tasks Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary 1.5.1

zEnsure that the system is not being used. It is advisable that the system stop accepting new LSF jobs while the archive is being made: % badmin hclos

1.5.2 Distribute Golden Image to a set of Compute Nodes To distribute a golden image to a set of Compute Nodes, you need to first register the image.

clsysfile creates an SD bundle CPACK-FILES. This bundle of files can be used to customize the files on the compute nodes. The revision number of the b

Note: It may take up to 24 hours to receive the license file. Plan accordingly. For more information, see the Comprehensive Instructions for this ste

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyTo install a CP

System Maintenance Tasks ClusterPack System Maintenance Tasks Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dicti

The steps in this section have to be followed in the specified order to ensure that everything works correctly. Step 1 Invoke /opt/clusterpack/bin/man

In the later case, the utility will prompt you (for each node in the cluster) whether to boot it or skip it. To boot a compute node with a system imag

zInstallation and configuration of the Management Server zInstallation and configuration of the Compute Nodes zVerification of the Management Server

Using CLI Software can also be installed on Compute Nodes using the /opt/clusterpack/bin/clsh tool to run the swinstall command. However, this may no

Using the CLI Software can also be removed from Compute Nodes using the /opt/clusterpack/bin/clsh tool to run the swremove command: zTo remove product

The process for updating software is the same as for installing software. (See "Install Software in Compute Nodes"). swinstall will verify t

account parameters to use in creating the account. If NIS is configured in the cluster, all user accounts are administered from the Management Server

account to remove. All user accounts are administered from the Management Server. Any changes to a users account will be pushed to all the Compute No

Step Q7 Configure the ProCurve Switch zSelect an IP address from the same IP subnet that will be used for the Compute Nodes. zConnect a console to

Back to Top 1.6.9 Define Compute Node Inventory Data Collection for Consistency checks Scheduling Data Collection tasks are done using the HP System

z3.2.3 How to Run SCM Web-based GUI Back to Top 1.6.10 Define Consistency Check Timetables on Compute Node Inventories Scheduling Data Collection ta

z3.2.3 How to Run SCM Web-based GUI Back to Top 1.6.11 Compare the Inventories of a Set of Nodes Comparing the results of Data Collection tasks is d

z3.2.3 How to Run SCM Web-based GUI Back to Top 1.6.12 Execute remote commands on one or more nodes A remote command can be executed on one or more

zUpdate /etc/checklist on node1, node3 and node5 with the local /etc/checklist % clcp -C node1+node3+node5 /etc/checklist %h:/etc/checklist zCopy mult

using PIDs on a cluster is not feasible given there will be different PIDs on different hosts, clkill can kill processes by name. Some examples of cl

Groups of Compute Nodes can be removed from ClusterPack using /opt/clusterpack/bin/clgroup. The following example removes the node group "cae&quo

Back to Top 1.6.20 Add File Systems to Compute Nodes The file system for Compute Nodes can be defined using System Administration Manager (SAM). Invo

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company% /sbin/init.d/

System Monitoring Tasks ClusterPack System Monitoring Tasks Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Diction

References: zStep 9 Install ClusterPack on the Management Server Back to Top Step Q10 Run manager_config on the Management Server Provide the follow

zState refers to the state of the host. zBatch State refers to the state of the host, and the state of the daemons running on that host. A detailed l

have exceeded their thresholds. zclosed_Excl - The host is not accepting jobs until the exclusive job running on it completes. zclosed_Full - The ho

or % bqueues -l <hostname> For more information, see the man page: % man bqueues Common Terms Both the Web interface and the CLI use the same te

Default status from each node is available using: % bhosts <hostname> STATUS shows the current status of the host and the SBD daemon. Batch jobs

1.7.4 View Usage of Resources in Compute Node(s) Using the Clusterware Pro V5.1 Web Interface: From the Hosts Tab: zSelect the host to be monitored u

z3.7.9 How do I access the Clusterware Pro V5.1 Command Line Interface? Back to Top 1.7.5 Monitor Compute Nodes based on resource thresholds Using t

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard company

Workload Management Tasks ClusterPack Workload Management Tasks Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dic

After adding, removing or modifying queues, it is necessary to reconfigure LSF to read the new queue information. This is done from the Management Ser

Back to Top 1.8.3 Restrict user access to specific queues Using the Clusterware Pro V5.1 CLI: The file /share/platform/clusterware/conf/lsbatch/<c

zStep 11 Run mp_register on the Management Server Back to Top Step Q12 Power up the Compute Nodes Use the clbootnodes program to power up all Compu

% lsid Find the queue definition you wish to modify. The following entries for maximum resource usage can be modified or added for each queue definiti

PRIORITY = <integer value> to the queue definition. Queues with higher priority values are searched first during scheduling. After adding, remov

to the queue definition. The command or tool should be accessible and runnable on all nodes that the queue services. After adding, removing or modifyi

Users can kill their own jobs. Queue administrators can kill jobs associated with a particular queue. References: z3.7.9 How do I access the Clusterwa

1.8.11 Suspend all jobs owned by a user Using the Clusterware Pro V5.1 CLI: All of a user's jobs can be suspended using the special 0 job id: %

Index | Administrators Guide | Users Guide | Tool Overview | Related Documents | Dictionary Copyright 1994-2004 hewlett-packard companyz3.7.9 How do I

System Troubleshooting Tasks ClusterPack System Troubleshooting Tasks Index | Administrators Guide | Users Guide | Tool Overview | Related Documents

% lshosts -l <hostname> % bhosts -l <hostname> References: z1.7.1 Get an Overview of Cluster Health z1.7.3 Get details on health of speci

This will reboot the machine, hostname, and cause the machine to install from the golden image you specified. References: z1.5.2 Distribute Golden Ima

Problem: When I try to add a node, I get "Properties file for <xyz> doesn't exist." Solution: zMake sure that the hostname is ful