Hp Insight Cluster Management Utility Manuale Utente Pagina 1

Navigare online o scaricare Manuale Utente per Software Hp Insight Cluster Management Utility. HP Insight Cluster Management Utility User Manual [en] Manuale Utente

  • Scaricare
  • Aggiungi ai miei manuali
  • Stampa
Vedere la pagina 0
HP Insight Cluster Management Utility v7.1
User Guide
Abstract
This guide describes how to install, configure, and use HP Insight Cluster Management Utility (CMU) v7.1 on HP systems. HP
Insight CMU is software dedicated to the administration of HPC and large Linux clusters. This guide is intended primarily for
administrators who install and manage a large collection of systems. This document assumes you have access to the documentation
that comes with the hardware platform where the HP Insight CMU cluster will be installed, and you are familiar with installing
and administering Linux operating systems.
HP Part Number: 5900-2346
Published: April 2013
Edition: 1
Vedere la pagina 0
1 2 3 4 5 6 ... 190 191

Sommario

Pagina 1 - User Guide

HP Insight Cluster Management Utility v7.1User GuideAbstractThis guide describes how to install, configure, and use HP Insight Cluster Management Util

Pagina 2

53 User group management...9954 Certificate error...

Pagina 3 - Contents

HP Insight CMU provides the latest conrep kit available at release time. If a different or newerversion of conrep is required for the servers in your

Pagina 4 - 4 Contents

1. In the /opt/cmu/etc/cmu_custom_menu file, uncomment the following line:SERVER;audit|dmidecode;/opt/cmu/bin/cmu_dsh -f CMU_TEMP_NODE_FILE -c "d

Pagina 5 - Contents 5

Help commandsTo get help during a CLI session, use the help command. This command displays all availablecommands of HP Insight CMU CLI.cmu> helpHEL

Pagina 6 - 6 Contents

halt nodes of logical group group_1 except node_exp halt delay "mesg" all group_1 group_2 halt nodes of group_1 an

Pagina 7 - Contents 7

Executing a command on a list of nodesTo execute a command on multiple nodes, you must specify the names of nodes.cmu> boot o185i222 o185i233 o185i

Pagina 8 - 8 Contents

Executing a command on specific nodes of a logical groupYou can use the but option to exclude active nodes of a group from the selection. Nodes to exc

Pagina 9

To broadcast on all nodes of the cluster:cmu> broadcast allselected nodes: o185i192 o185i193 o185i194 o185i195 o185i196 o185i197 o185i198 o185i199

Pagina 10 - Examples

active node list selected: o185i192Please read /opt/cmu/log/PowerOff.log for errors.cmu>Setting the locator LED on or offSets the locator LED of a

Pagina 11 - 1 Overview

Total | 1 | 0 | 0Detailed logs are in /opt/cmu/log/cmucerbere.log and/opt/cmu/log/cmucerbere-*.log

Pagina 12 - 1.1.4 System disk replication

[16:15:13] OSTYPE:Linux-CMU[16:15:13] [DollyClient] Starting to get fstab files[16:15:13] [DollyClient] Getting "/opt/cmu/tmp/fstab.txt"[16:

Pagina 13 - 2.1 Installing HP Insight CMU

1 OverviewHP Insight Cluster Management Utility (CMU) is a collection of tools that manage and monitor alarge group of computer nodes, specifically HP

Pagina 14 - 2.1.3 Disk space requirements

[16:25:06] [DollyClient] Device is sda[16:25:06] [DollyClient] Asking for partition table of "/dev/sda"[16:25:06] [DollyClient] Getting /opt

Pagina 15

6.17.5 Administration utilities pdcp and pdshHP Insight CMU includes the open source software pdcp and pdsh.Usage example of pdcp:# /opt/cmu/bin/pdcp

Pagina 16 - ◦ Configure SATA as IDE

7 Advanced topics7.1 Accessing the GUI for non-root usersHP Insight CMU allows non-root users to log into the GUI and access some or all of the privil

Pagina 17 - 2.1.7.3 DL160 G6 Servers

Table 3 Operational HP Insight CMU GUI features available by default for non-root users (continued)user (requires sudo)Cloning (Deploy Image)user (req

Pagina 18

Table 4 HP Insight CMU GUI features and their corresponding commandsHP Insight CMU management node commandHP Insight CMU GUI feature (right-click node

Pagina 19

In this context, the term "diskless" refers to any OS image that can be created and prepared locallyon the HP Insight CMU management server

Pagina 20 - 2.2.3.1 RHEL 6 support

-l <CMU diskless logical group name>The name of the logical group to delete.The delete_image program is expected to delete everything related to

Pagina 21 - 2.2.6 Login privileges

-n <nodename>The hostname of the target node to boot.-i <IP address>The IP address of the target node to boot.-m <MAC address>The MA

Pagina 22 - 2.3 Installation procedures

ILOCMThe method for integration with HP Moonshot 1500 Chassis.The HP Insight CMU hardware API consists of a collection of programs that reside in /opt

Pagina 23

CMU_VALID_HARDWARE_TYPES=ILO:lo100i:ILOCMTo add the IPMI hardware API, add IPMI to the list of valid hardware types:CMU_VALID_HARDWARE_TYPES=ILO:lo100

Pagina 24

• Managing the system images stored by HP Insight CMU• Configuring actions performed when a node status changes such as display a warning, executea co

Pagina 25 - 2.4.2 Software prerequisites

etc/bootopts/AC14000. The hexadecimal IP address AC14000 covers IP addresses 172.20.0.1- 172.20.0.15.7.5 Support for ScaleMPHP Insight CMU can be inte

Pagina 26

The transfer uses TCP/IP sockets. The clone image is saved to the local disk. The node then asksthe image server if any successors are waiting for upl

Pagina 27

122 Advanced topics

Pagina 28

8 Support and other resources8.1 Contacting HP8.1.1 Before you contact HPBe sure to have the following information available before you contact HP:• T

Pagina 29

• Installation and user guides for your specific operating system.8.3 Typographic conventionsThis document uses the following typographical convention

Pagina 30 - 2.5 Upgrading HP Insight CMU

CAUTIONA caution calls attention to important information that if not understood or followed will resultin data loss, data corruption, or damage to ha

Pagina 31 - 2.5.7 Starting HP Insight CMU

A TroubleshootingIssues encountered while using HP Insight CMU can be classified as:• Network boot issues which affect cloning and backup• Backup spec

Pagina 32

• An incorrect MAC address in the HP Insight CMU database• The HP Insight CMU configuration on the management node is lost.Troubleshooting switch issu

Pagina 33 - 3.2.3 Administrator mode

A.4 Cloning issuesIf only one node cannot be cloned:1. Verify that you can boot in network mode.2. Verify that the node has the same hardware as other

Pagina 34 - 3.4 Cluster administration

3. Verify that rsh or ssh is enabled between all nodes of the cluster and the management node.All nodes must be able to execute commands as root for a

Pagina 35 - 3.4.1 Node management

2 Installing and upgrading HP Insight CMU2.1 Installing HP Insight CMUA typical HP Insight CMU cluster contains three kinds of nodes. Figure 1 (page 1

Pagina 36 - 3.4.1.1 Scanning nodes

On Windows, go to System Preferences→Other→Java→Advanced→Enable online certificatevalidation. On Linux, run javaws -viewer in a shell, click the Advan

Pagina 37 - 3.4.1.2 Adding nodes

B Detailed installation instructionsB.1 Install required RPMs1. Install expect library.2. Install DHCP.3. Install the TFTP server.4. Install the TFTP

Pagina 38 - 3.4.1.3 Modifying nodes

• On SLES:# chkconfig nfsserver on# /etc/init.d/nfsserver startB.4 Verifying the DHCPD listen interfaceVerify that DHCPD is correctly configured to li

Pagina 39 - 3.4.1.7 Contextual menu

3. Install the HP Insight CMU rpm:# rpm --import /mnt/cmuteam-rpm-key.asc# rpm -ivh /mnt/cmu-v7.1-1.i386.rpmPreparing... ##############

Pagina 40

1. Edit the /opt/cmu/etc/cmuserver.conf file:# vi /opt/cmu/etc/cmuserver.conf2. Search for the CMU_CLUSTER_IP variable.3. Replace the default value wi

Pagina 41 - 4.1 Logical group management

monitoringStatus of the monitoring daemon that gathers the information reported by the small monitoringagent installed on the compute nodes.web servic

Pagina 42 - 4.1.3 Renaming logical groups

B.14.1 Configuring the GUI client on Linux workstationsOn Linux workstations, you can use a secure ssh tunnel or an X Window server to communicatebetw

Pagina 43 - 4.2 Autoinstall

• The server access control must allow access. To authorize access, use the xhost + command.• Allow rmi connection and X display export in your firewa

Pagina 44 - 4.2.4.1 Enabling autoinstall

Figure 56 HP Insight CMU GUINOTE: At this point in the installation process, the GUI window will not contain most of the detailsshown in the previous

Pagina 45 - 4.2 Autoinstall 45

HP Insight CMU manpages139

Pagina 46

2.1.2 Planning for compute node installationTwo IP addresses are required for each compute node.• Determine the IP address for the management card (iL

Pagina 47 - 4.2.6 Customization

cmu_show_nodes(8)NAMEcmu_show_nodes -- Display a list of nodes and node attributes.SYNOPSIS# /opt/cmu/bin/cmu_show_nodes [-a | -n <node>] [-i] [

Pagina 48 - 4.3 Backing up

%c(ILOCM only) cartridge number%N(ILOCM only) node numberEXAMPLESDefault behavior:# /opt/cmu/bin/cmu_show_nodescn0004cn0005cn0006cn0008cn0009To show d

Pagina 49 - 4.3 Backing up 49

cmu_show_logical_groups(8)NAMEcmu_show_logical_groups -- Show nodes belonging to a logical group.SYNOPSIS# /opt/cmu/bin/cmu_show_logical_groups <-h

Pagina 50 - 4.4 Cloning

cmu_show_network_entities(8)NAMEcmu_show_network_entities -- Show network entities.SYNOPSIS# /opt/cmu/bin/cmu_show_network_entities <-h | [network_

Pagina 51 - 4.4.1 Preconfiguration

cmu_show_user_groups(8)NAMEcmu_show_user_groups -- Show user groups.SYNOPSIS# /opt/cmu/bin/cmu_show_user_groups <-h | [user_group]>DESCRIPTIONSh

Pagina 52 - 4.4.2 Reconfiguration

cmu_show_archived_user_groups(8)NAMEcmu_show_archived_user_groups -- Show archived user groups.SYNOPSIS# /opt/cmu/bin/cmu_show_archived_user_groups [-

Pagina 53 - 4.6 Rescan MAC

cmu_add_node(8)NAMEcmu_add_node -- Add node(s) to the HP Insight CMU database.SYNOPSIS# /opt/cmu/bin/cmu_add_node <-h | -s | -i | -f filename>#

Pagina 54 - 4.7.1 Expanding an image

EXAMPLESCommand-line mode:# /opt/cmu/bin/cmu_add_node -H cn0006 -I 16.16.184.116 -M 255.255.254.0 -A 00-02-A5-52-EB-F8 -L default -G 192.168.0.1 -T IL

Pagina 55 - 4.8.1 Overview

cmu_add_network_entity(8)NAMEcmu_add_network_entity -- Add network entities.SYNOPSIS# /opt/cmu/bin/cmu_add_network_entity <-f filename | -h># /o

Pagina 56 - On the golden node

cmu_add_logical_group(8)NAMEcmu_add_logical_group -- Add logical groups.SYNOPSIS# /opt/cmu/bin/cmu_add_logical_group <-n | -i | -f filename | -s>

Pagina 57 - From the GUI

NOTE: On Blade servers, to configure the IP addresses on the iLO cards, you can use theEBIPA on the OA. For instructions, see “Configuring iLO cards f

Pagina 58

cmu_add_to_logical_group_candidates(8)NAMEcmu_add_to_logical_group_candidates -- Add nodes as candidates for logical groups.SYNOPSIS# /opt/cmu/bin/cmu

Pagina 59 - From the CLI

cmu_add_user_group(8)NAMEcmu_add_user_group -- Add user groups.SYNOPSIS# /opt/cmu/bin/cmu_add_user_group <-f filename | -h># /opt/cmu/bin/cmu_ad

Pagina 60 - 4.8.12.1 files.custom

cmu_add_to_user_group(8)NAMEcmu_add_to_user_group -- Add nodes to user groups.SYNOPSIS# /opt/cmu/bin/cmu_add_to_user_group <-h | -t user_group node

Pagina 61

cmu_change_active_logical_group(8)NAMEcmu_change_active_logical_group -- Change the active logical group for a node.SYNOPSIS# /opt/cmu/bin/cmu_change_

Pagina 62

cmu_change_network_entity(8)NAMEcmu_change_network_entity -- Change the network entity for a node.SYNOPSIS# /opt/cmu/bin/cmu_change_network_entity <

Pagina 63 - On Red Hat

cmu_del_from_logical_group_candidates(8)NAMEcmu_del_from_logical_group_candidates -- Delete nodes from logical groups.SYNOPSIS# /opt/cmu/bin/cmu_del_f

Pagina 64

cmu_del_from_network_entity(8)NAMEcmu_del_from_network_entity -- Delete nodes from network entities.SYNOPSIS# /opt/cmu/bin/cmu_del_from_network_entity

Pagina 65

cmu_del_archived_user_group(8)NAMEcmu_del_archived_user_group -- Delete an archived user group.SYNOPSIS# /opt/cmu/bin/cmu_del_archived_user_group [-h]

Pagina 66

cmu_del_from_user_group(8)NAMEcmu_del_from_user_group -- Delete one or more nodes from a user group.SYNOPSIS# /opt/cmu/bin/cmu_del_from_user_group <

Pagina 67 - 5.3 Monitoring the cluster

cmu_del_logical_group(8)NAMEcmu_del_logical_group -- Delete a logical group.SYNOPSIS# /opt/cmu/bin/cmu_del_logical_group <-f filename | -h># /op

Pagina 68 - 5.3.1 Node and group status

2.1.7.1.2 Configuring iLO cards from the OA: Blades onlyUse the EBIPA to assign consecutive addresses to the iLO:• 16 addresses on the c7000 Enclosure

Pagina 69 - 5.3 Monitoring the cluster 69

cmu_del_network_entity(8)NAMEcmu_del_network_entity -- Delete a network entity.SYNOPSIS# /opt/cmu/bin/cmu_del_network_entity <-f filename | -h>#

Pagina 70

cmu_del_node(8)NAMEcmu_del_node -- Delete a node.SYNOPSIS# /opt/cmu/bin/cmu_del_node <-f filename | -h># /opt/cmu/bin/cmu_del_node <node_name

Pagina 71 - 5.3.5 Gauge widget

cmu_del_snapshots(8)NAMEcmu_del_snapshots -- Delete monitoring snapshots from the history database.SYNOPSIS# /opt/cmu/bin/cmu_del_snapshots [-h] | <

Pagina 72 - 5.3.7 Using time view

cmu_del_user_group(8)NAMEcmu_del_user_group -- Delete a user group.SYNOPSIS# /opt/cmu/bin/cmu_del_user_group <-f filename | -h> [-a] [-m]# /opt/

Pagina 73 - ◦ Launch HP Insight CMU:

cmu_console(8)NAMEcmu_console -- Connect to compute node management ports.SYNOPSIS# /opt/cmu/bin/cmu_console <compute_node_hostname>DESCRIPTIONI

Pagina 74 - 5.3.7.4 Bindings and options

cmu_power(8)NAMEcmu_power -- Perform power actions on compute nodes.SYNOPSIS# /opt/cmu/bin/cmu_power <-h | -p action -n nodename1 [nodename2] [node

Pagina 75 - 5.3.7.6 Troubleshooting

EXAMPLESTo power off one node:.cmu_power -p OFF -n cn0001To power off nodes belonging to user group user1:.cmu_power -p OFF -u user1To boot nodes belo

Pagina 76 - 5.3.8 Archiving user groups

cmu_custom_run(8)NAMEcmu_custom_run -- A CLI to HP Insight CMU custom menu options.SYNOPSIS# /opt/cmu/bin/cmu_custom_run <-h | -l | -t command_titl

Pagina 77 - 5.5.1 Action and alert files

cmu_clone(8)NAMEcmu_clone -- Clone nodes in a logical group.SYNOPSIS# /opt/cmu/bin/cmu_clone <-n | -f nodelistfile> <-i imagename> [-s sum

Pagina 78 - 5.5.2 Actions

cmu_backup(8)NAMEcmu_backup -- Issue backup commands directly from the Linux shell.SYNOPSIS# /opt/cmu/bin/cmu_backup <-h> | <-l logical_group

Pagina 79 - 5.5.4 Alert reactions

NOTE: These IDE settings only apply to the DL160 G5 Server.• IPMISerial Port assigned to System◦◦ Serial Port Switching Disabled◦ Serial Port Connecti

Pagina 80

cmu_scan_macs(8)NAMEcmu_scan_macs -- Scan IP addresses and create HP Insight CMU node definitions.SYNOPSIS# /opt/cmu/bin/cmu_scan_macs -h <hostname

Pagina 81

when there is an intervening empty slot. The -S 0 option effectively forces a sequential set ofvalues to be generated for %xi and the IP since interve

Pagina 82 - + (cputotals.sys)

EXAMPLESExample 1To scan 128 sequential ILO addresses starting at 3.4.5.6 and put node definitions similar to thefollowing in the HP Insight CMU datab

Pagina 83

n03_C01_N3 1.2.3.3 255.255.0.0 44-1e-a1-d3-b4-02 default 10.84.202.42 ILOCM x86_64 1 3n04_C01_N4 1.2.3.4 255.255.0.0 44-1e-a1-d3-b3-de default 10.84.2

Pagina 84

cmu_rescan_mac(8)NAMEcmu_rescan_mac -- Rescan the MAC address of a node.SYNOPSIS# /opt/cmu/tools/cmu_rescan_mac -n nodename [N NIC_num] [-h]DESCRIPTIO

Pagina 85

cmu_mod_node(8)NAMEcmu_mod_node -- Add node(s) to the HP Insight CMU database.SYNOPSIS# /opt/cmu/bin/cmu_mod_node <-h | -s | -i | -f filename>#

Pagina 86 - 5.5.7.2 Monitoring AMD GPUs

# /opt/cmu/bin/cmu_mod_node -H cn0006 -I 16.16.184.116 -M 255.255.254.0-A 00-02-A5-52-EB-F8 -L default -G 192.168.0.1 -R x86_64processing 1 node ...In

Pagina 87

cmu_monstat(8)NAMEcmu_monstat -- Use monitoring to list sensors and alerts.SYNOPSIS# /opt/cmu/bin/cmu_monstat <--alerts=alert1 | --all-alerts | --a

Pagina 88

--all-lgSelect all logical groups.--all-neSelect all network entities--all-ugSelect all user groups--lg=lg1,lg2,...Specify the logical group(s) names

Pagina 89 - 5.5.9 Extended metric support

cmu_image_open(8)NAMEcmu_image_open -- Open an existing backup image for modification.SYNOPSIS# /opt/cmu/bin/cmu_image_open <-h | -i imagename>D

Pagina 90

2.1.7.4 SL2x170z G6 and DL170h G6 Servers BIOS settingIMPORTANT: To enable BIOS updates, you must restart the server. You can restart the serverwith C

Pagina 91 - 6.3 SSH connection

cmu_image_commit(8)NAMEcmu_image_commit -- Save a backup image previously expanded with cmu_image_open.SYNOPSIS# /opt/cmu/bin/cmu_image_commit <-h

Pagina 92 - 6.7 Power off

cmu_config_nvidia(8)NAMEcmu_config_nvidia -- Configure NVIDIA GPU monitoring.SYNOPSIS# /opt/cmu/bin/cmu_config_nvidia <-h | -r | -n numGPUs>Wher

Pagina 93 - 6.10 Change UID LED status

cmu_config_amd(8)NAMEcmu_config_amd -- Configure AMD GPU monitoring.SYNOPSIS# /opt/cmu/bin/cmu_config_amd <-h | -n numGPUs>Where numGPUs specifi

Pagina 94 - 6.12 Single window pdsh

cmu_config_intel(8)NAMEcmu_config_intel -- Configure Intel coprocessor monitoring.SYNOPSIS# /opt/cmu/bin/cmu_config_intel <-h | -r | -n>DESCRIPT

Pagina 95 - 6.12.1 cmudiff examples

cmu_mgt_config(8)NAMEcmu_mgt_config -- Configure or test a set of Linux components required by HP Insight CMU.SYNOPSIS# /opt/cmu/bin/cmu_mgt_config [-

Pagina 96

ssh_keyCheck for existence of the root ssh key or create one.firewallCheck and optionally disable the firewall.tftpCheck and configure tftp.nfsCheck a

Pagina 97 - ConnectTimeout 1

cmu_firmware_mgmt(8)NAMEcmu_firmware_mgmt -- Verify and execute firmwareSYNOPSIS# /opt/cmu/bin/cmu_firmware_mgmt [-h] [-d -f <nodefile>[-o"

Pagina 98 - 6.14 User group management

Glossaryadministration disk The disk located on the image server on which HP Insight CMU is installed. A dedicated spacecan be allocated to the cloned

Pagina 99 - 6.14.3 Renaming user groups

2. A software package that is capable of being installed or removed with the RPM softwarepackage management.secondary server A dedicated node in a net

Pagina 100 - 6.16 Customizing the GUI menu

IndexAaction files, 78actionsandalerts.txt, 81adding network entities, 40adding nodes, 37adding user groups, 98administration, 12cluster, 34administra

Pagina 101 - 6.17 HP Insight CMU CLI

Otherwise, if your node is wired with a dedicated management port for LO100i:◦ BMC NIC Allocation Dedicated◦ LAN protocol: HTTP, telnet, ping Enabled•

Pagina 102 - Getting help for a command

Eextended metrics, 89Ffirewall, 132firmwareinstalling, 100upgrading, 100firmware management, 99firmware requirements, 14Gglossary, 187group status, 68

Pagina 103 - 6.17.3 Specifying nodes

NVIDIA GPUs, 85Ooperating system support, 20Pparametersexamples, 15pdcp, 97, 111pdsh, 94, 111power off, 92preconfiguration, 51provisioning, 41RRAID co

Pagina 104

© Copyright 2013 Hewlett-Packard Development Company, L.P.Confidential computer software. Valid license from HP required for possession, use or copyin

Pagina 105 - Booting a set of nodes

2.2.3 Operating system supportHP Insight CMU software is generally supported on Red Hat Enterprise Linux (RHEL) 5 and 6; andSUSE Linux Enterprise Serv

Pagina 106 - Rebooting a set of nodes

Table 1 Directory structure (continued)ContentsSubdirectoryDocumentation and release notesDocumentationContains the following licenses: Apache_LICENSE

Pagina 107 - Cloning a set of nodes

2.3 Installation procedures1. Perform a full installation of your base OS on the management node.2. HP Insight CMU depends on Oracle Java version 1.6

Pagina 108 - Backing up a node

9. Install HP Insight CMU on the GUI client workstation. For details, see “Installing HP InsightCMU on the GUI client workstation” (page 135).2.4 Inst

Pagina 109 - 6.17 HP Insight CMU CLI 109

The next figure shows a “classic” HP Insight CMU cluster with one HP Insight CMU managementserver and compute nodes connected directly to the site net

Pagina 110

2.4.1 HA hardware requirementsThe hardware requirements for HP Insight CMU under HA control are:• Two or more management servers.• One shared storage

Pagina 111 - 6.17 HP Insight CMU CLI 111

2.4.3.2 HP Insight CMU HA service requirementsWhen you configure the HA software layer, configure the HP Insight CMU HA service with thefollowing reso

Pagina 112 - 7 Advanced topics

* it must support locking via flock() ** it must be mounted only by one (active) cmu mgt node at a time ** it must

Pagina 113

cmu ha:cmu service needs (re)startThis command does not actually start HP Insight CMU. It only clears the audit mode to enableHP Insight CMU to be sta

Pagina 114 - 7.1.3 Examples

cmuadmin1cmuadmin2e. Unset the audit mode on the new member:# /etc/init.d/cmu unset_auditcmu ha:cmu service needs (re)startf. Start HP Insight CMU und

Pagina 115 - 7.2.2 Delete diskless image

Contents1 Overview...111.1 Features...

Pagina 116 - 7.2.5 Boot diskless node

12. Restore the cluster-wide configuration on server 1.13. Unset the audit mode on server 1.14. Using the appropriate command for your HA software, re

Pagina 117 - 7.2.6 Diskless check

2.5.5 Installing the HP Insight CMU v7.1 packageFor more information about installing the HP Insight CMU v7.1 package, see “Installationprocedures” (p

Pagina 118 - 118 Advanced topics

3 Defining a cluster with HP Insight CMU3.1 HP Insight CMU service statusObtain the status of all HP Insight CMU service components with the following

Pagina 119

Figure 4 (page 32) contains four main areas:• The top bar allows you to perform configuration commands.• The left frame lists resources such as Networ

Pagina 120 - 7.6 Cloning mechanisms

NOTE: If the Display Number field is empty, verify that you started your X server and that yourfirewall allows X traffic.3.3 High-level checklist for

Pagina 121 - 7.6 Cloning mechanisms 121

3.4.1 Node managementFigure 7 Node management windowIn Figure 7 (page 35), the node list of the cluster will appear as the node database is populatedb

Pagina 122 - 122 Advanced topics

3.4.1.1 Scanning nodesCluster Administration→Node Management→Scan NodeThe HP Insight CMU Node Management component provides the capability to scan new

Pagina 123 - 8 Support and other resources

NOTE: This is necessary only for the first scan operation. For subsequent scans, theManagement card password window will not be displayed.Figure 9 Man

Pagina 124 - 8.3 Typographic conventions

Figure 11 Add node dialogAt the Node Dialog box:1. Click OK. A dialog box displays the successful addition of a node completion.2. Click OK. A dialog

Pagina 125

To modify the attributes of a node, select the node in the Node Management list, and then selectModify Node. The same interface as Add Node appears.NO

Pagina 126 - A Troubleshooting

2.5.5 Installing the HP Insight CMU v7.1 package...312.5.6 Restoring the HP Insight CMU

Pagina 127 - A.3 Backup issues

You can use the Network Entity Management window to add and delete network entities. Toperform tasks by using the Network Entity Management option, cl

Pagina 128 - A.6 GUI problems

4 Provisioning a cluster with HP Insight CMU4.1 Logical group managementA logical group in HP Insight CMU represents a disk image that has been captur

Pagina 129 - A.6 GUI problems 129

• For the first smart array logical drive on ProLiant servers, use cciss/c0d0.IMPORTANT: For RHEL6, the smart array device name depends on the smart a

Pagina 130 - 130 Troubleshooting

4.2 AutoinstallThe HP Insight CMU kickstart functionality is renamed autoinstall. HP Insight CMU autoinstallprovides the following improvements:• Adds

Pagina 131 - B.1 Install required RPMs

4.2.4 Using autoinstall from GUI4.2.4.1 Enabling autoinstallBy default, the HP Insight CMU GUI does not display the autoinstall buttons. To enable thi

Pagina 132 - B.7 Installing HP Insight CMU

Figure 18 New autoinstall logical groupAfter the autoinstall logical group is created, the HP Insight CMU image directory contains a newdirectory with

Pagina 133 - B.9 Setting the Java PATH

NOTE: Autoinstall files and pxelinux files are created only if they do not already exist. Thisenables parameters to be customized for a node or group

Pagina 134 - B.11 Starting HP Insight CMU

cmu> add_to_logical_group node1 to rh5u5_autoinstselected nodes: node1 processing 1 node ... cmu>Or:# /opt/cmu/bin/cmu_add_to_logical_group_c

Pagina 135

4.2.7 RestrictionsThis implementation contains the following restrictions:• The repository must be on the local storage of the management node.• The r

Pagina 136 - Using an X Window server

IMPORTANT: If partitions to be backed up are less than 50% empty, you must configure HPInsight CMU to use the tmpfs file system for cloning partitions

Pagina 137

4.6 Rescan MAC...534.7 HP Insight CMU

Pagina 138 - Figure 56 HP Insight CMU GUI

4.4 CloningThe HP Insight CMU cloning operation copies the complete contents of the golden image to othernodes. The copied image is the same except fo

Pagina 139 - HP Insight CMU manpages

Figure 23 Cloning statusWhen cloning is complete, a popup window displays the results.The correctly cloned compute nodes appear in the chosen logical

Pagina 140 - DESCRIPTION

The default content of pre_reconf.sh is:#!/bin/bash#keep this version tag hereCMU_PRE_RECONF_VERSION=1#starting from cmu version 4.2 this script is de

Pagina 141 - EXAMPLES

# CMU_RCFG_IP = mgt network ip of this compute node# CMU_RCFG_NTMSK = net maskexit 04.5 Node static infoTo collect static information such as system m

Pagina 142

Figure 25 Rescan MAC4.7 HP Insight CMU image editorAn existing HP Insight CMU cloning image can be modified directly on the HP Insight CMUmanagement n

Pagina 143

4.7.2 Modifying an imageModifications can consist of simple manual commands such as adding, removing, or modifyingfiles. However, complex operations u

Pagina 144

In the HP Insight CMU implementation, the compute nodes share the operating system on the HPInsight CMU management node. Each compute node has its own

Pagina 145

user = root server = /usr/sbin/in.tftpd server_args = /tftpboot /opt/cmu/ntbt/tftp -v

Pagina 146

Figure 26 Adding a new logical group3. Select the Diskless option to the right of the group name.NOTE: If you cannot see the Diskless option, the disk

Pagina 147

7. Select one of these kernels, and then click OK. The diskless image building process launches.This operation might last several minutes while files

Pagina 148

5.5.2 Actions...785.5.3 Alerts...

Pagina 149

4.8.10 Booting the compute nodesFrom the GUI1. Select the compute nodes you added to the diskless logical group.2. Right-click to launch a boot comman

Pagina 150

4.8.12.2 Using reconf-diskless-image.shThe reconf-diskless-image.sh script is executed at the end of the image building process.This script contains a

Pagina 151

#!/bin/bash#cmu_begin_interface#do not change anything in this section#add custom code after this sectionCMU_RECONF_DISKLESS_SNAPSHOT_VERSION=1# start

Pagina 152

◦ The snapshot directories are not synchronized. The registration process copies the listedfiles into files and files.custom in the snapshot directory

Pagina 153

On SLES# chkconfig nfsserver on3. Ensure that enough NFS daemons and threads are configured to handle the anticipated volumeof NFS traffic.On Red HatS

Pagina 154

When a node is added to the diskless logical group• A copy of the snapshot directory for this node is sent to the NFS server.• A PXE-boot file is crea

Pagina 155

5 Monitoring a cluster with HP Insight CMU5.1 Installing the HP Insight CMU monitoring clientYou must install the HP Insight CMU monitoring client to

Pagina 156

5.3 Monitoring the clusterLaunch the HP Insight CMU GUI.Figure 31 Main windowIn Figure 31 (page 67), the left frame lists the resources, such as Netwo

Pagina 157

Figure 32 Node statusThe status of this node is okay. Node values are correctly reported to the main monitoring daemon.The node is pinging properly, a

Pagina 158

In the central frame, the following tabs are available:• Instant View• Table View• Time View• Details• AlertsFor a single node view, the following tab

Pagina 159

7.2.2 Delete diskless image...1157.2.3 Configure diskless

Pagina 160

5.3.4 Resource view in the central frameMonitoring values can be visualized by:• Global cluster• A specific logical group• A specific network entity•

Pagina 161

5.3.4.2 Detail mode in resource viewTo display a table with sensor values, select the Instant View tab in the central frame.• The cell is green when t

Pagina 162

• Details — Shows static data for the node. Some of the values are filled during the initial nodediscovery (scan node). Other values are filled by rig

Pagina 163

5.3.7.1 Getting startedTo launch HP Insight CMU with Time View:• From the web:Go to http://yourcluster. Click the first link Launch Insight Cluster Ma

Pagina 164

Figure 39 Time view5.3.7.4 Bindings and options5.3.7.4.1 Mouse control• Left-click on a node – Mark the node from a set of four predefined colors• Rig

Pagina 165

5.3.7.4.3 Custom camerasTo save a custom camera position, press Ctrl+1 to 5. Restore it later by pressing 1 to 5. (Customcamera position 1 ... 5 optio

Pagina 166

Some GPUs may not support anti-aliasing levels set to 8. Symptoms are black strips on the left andright of Time View, or cylinders above the rings mak

Pagina 167

5.3.8.2 LimitationsTo display an archived user group, the following conditions must be satisfied:• Time must not exceed 24 hours.• The number of nodes

Pagina 168

### ALERTS###cpu_freq_alert "CPU frequency is not nominal" 1 24 100 < % sh -c "b=`cat /sys/devices/syste

Pagina 169

• MeanOverTime returns the difference between the current value and the previous valuedivided by the time interval.For example, if the sensors return

Pagina 170 - OPTIONS (naming)

cmu_add_network_entity(8)...148cmu_add_logical_group(8

Pagina 171 - OPTIONS (general)

ConditionThe reaction is performed under this condition.• ReactOnRaise — Execute the reaction whenever the alert shows as raised and the previousstate

Pagina 172 - Example 3

• Add your own sensors, alerts, or alert reactions by adding a line to the ACTIONS, ALERTS,or ALERT_REACTIONS section.Modifications in the ActionAndAl

Pagina 173 - Example 4

#- Native#cpuload "% cpu load (raw)"1 numerical MeanOverTime 100 % awk '/cpu / {printf"%d\n",$2+$3+$4}' /proc/stat#- Co

Pagina 174

For more information about using and fine tuning collectl, see http://collectl.sourceforge.net/.5.5.6.3 Installing and configuring colplot for plottin

Pagina 175

9. Import the common directory created on the administration server for collectl.# mkdir /var/log/collectl# vi /etc/fstabX.X.X.X:/var/log/collectl /

Pagina 176

Select plotting options, then click Generate Plot.Figure 43 ColPlot results5.5.7 Monitoring GPUs and coprocessors5.5.7.1 Monitoring NVIDIA GPUsIf your

Pagina 177 - NODE AND GROUP OPTIONS

..Running /opt/cmu/bin/cmu_config_nvidia adds a list of predefined GPU metrics toActionAndAlertsFile.txt. To monitor these metrics using the GUI, sele

Pagina 178

5.5.7.3 Monitoring Intel coprocessorsIf your client nodes contain Intel coprocessors, you can monitor the coprocessors with HP InsightCMU.Install the

Pagina 179

k. Review the results and verify no errors are reported.l. With the coprocessors working, enable coprocessor monitoring by updating the /opt/cmu/etc/A

Pagina 180

keywords such as CMU_ALERT_NODES can be used to convey the names of the nodes that raisedthe alert through the SNMP trap.Figure 44 HP Insight CMU aler

Pagina 181

Figures1 Typical HPC cluster...132 iLO server

Pagina 182

data is received after this time interval expires, the GUI marks the extended metric data"invalid".Data TypeA description of the format of t

Pagina 183

6 Managing a cluster with HP Insight CMUCluster management tasks can be performed on one or more nodes with HP Insight CMU. Thesetasks depend on your

Pagina 184

To select a terminal emulator other than the default:1. Edit /opt/cmu/etc/cmuserver.conf.2. Six blocks of variable names begin with CMU_REMOTE_TERMINA

Pagina 185

Figure 47 Power off dialog box6.8 BootWhen one or more nodes are selected, this task enables you to boot a collection of nodes on theirown local disk

Pagina 186

6.11 Multiple windows broadcastThis task is available when one or more nodes are selected. The following connections are availablefor multiple windows

Pagina 187 - Glossary

Figure 51 pdsh windowYou can toggle the two filters on and off using dshbak or cmudiff. These two filters are mutuallyexclusive, so you can:• Filter w

Pagina 188 - 188 Glossary

• Some details about output processing results, which are provided on the right.Characters that differ from the reference node are highlighted in red.

Pagina 189

cmudiff filter is <ON>, with parameters -d cmu_pdsh>cmu_pdsh> dmidecodeThe comment now shows “(2 populations) o185i[040,042] are 83% simi

Pagina 190 - 190 Index

Figure 52 Parallel distributed copy window3. Complete the Source and Destination fields, and then click OK to execute the distributed copy.6.14 User g

Pagina 191

Figure 53 User group managementSelect any number of nodes from the list of “Nodes in Cluster” on the left and use the arrows tomove the nodes to the l

Commenti su questo manuale

Nessun commento