Manuals / Brands / Computer Equipment / Software / HP / Computer Equipment / Software

HP UX 11i v2 Networking Software 7 Monitoring and Troubleshooting, Diagnosing Problems

1 236

Download 236 pages, 9.65 Mb

7 Monitoring and Troubleshooting

This chapter discusses monitoring and troubleshooting your HP Fabric Clustering System. The HP-UX Host and HP Fabric Switch each have a unique set of tools for monitoring and troubleshooting their respective components within the HP Fabric Clustering System. This chapter contains the following sections:

•“Diagnosing Problems”

•“Monitoring and Troubleshooting the HP-UX Host”

•“Monitoring and Troubleshooting the Switch”

Diagnosing Problems

This section provides useful guidelines to determine which component within your HP Fabric Clustering System is at fault. Follow the LED descriptions to determine the faulty component, and then proceed to the appropriate component troubleshooting section.

The power supply, system, and fan LEDs are located on the front of the HP Fabric Switch. The back contains HP Fabric Cluster Connection LEDs, Ethernet Management Port LEDs, and duplicated System Status LED.

•“LED Color, Behavior, and Meaning”

—“HP Fabric Switch Cluster Connection LEDs”

—“HP Fabric Switch System Status”

—“Power Supply Troubleshooting”

—“HP Fabric HCA Connection LEDs”

—“Ethernet Management Port LEDs”

•“Determining Whether the HCA or Cable is Faulty”

•“Determining if the Switch is Faulty”

•“Next Steps”

Diagnosing Problems 109

Contents

HP-UXIntegrity Server Networking Page Table of Contents About This Document 1 Introduction to Technology 2 Hardware Overview 3 Installation Planning 5 Configuration 6 Administration and Management Using Command Completion Specifying the Card/Port Testing Network Connectivity Logging Onto the System Logging On Through SNMP 7 Monitoring and Troubleshooting A Specifications B Switch Command Line Interface link-trap C How to Use Windows HyperTerminal Glossary Index Page List of Figures Page List of Tables B-38 B-39 B-40 B-41 B-42 About This Document Typographical Conventions Related Documents HP Encourages Your Comments Page 1 Introduction to Technology HP Fabric Clustering System What is InfiniBand InfiniBand Advantages InfiniBand Capabilities InfiniBand Transport Services HP Fabric Subnet Manager MPI IPoIB Oracle 10gRAC User-SpaceRDMA Library HCA Kernel RDMA Subsystem Interaction with HyperFabric Driver Page 2 Hardware Overview Host Channel Adapters •AH304A PCIe adapter with 2 (two) 4X fabric copper interfaces •AD313A PCIe adapter with 2 (two) 4X fabric copper interfaces •AB286C PCI-X133Mhz adapter with 2 (two) 4X fabric copper interfaces Page Page Page Page Host Adapter View) Figureigure Page Switch Other Product Elements Page 3 Installation Planning Application Availability 4 Installing HP Fabric Clustering System Installing the AH304A Host Channel Adapter Installing the 410533-B21Host Channel Adapter Installing the AD313A Host Channel Adapter Installing the AB286C Host Channel Adapter On-LineAddition and Replacement Operations (OL*) Replacing an HP Fabric Clustering System HCA (HP-UX11i v2 only) Install HP Fabric Clustering System Software 1.Log on to the system as root The drivers, utilities and manpages for the software are on the •Go to http://www.software.hp.com Go to Rack Mount Preparation Installing the Switch Page Remove Shipping Sleeve Remove Support Bracket (if installed) Cable Routing and Bend Limits Attach Cables Switch or Adapter Attach to other HP Fabric devices: Connect Cables to Switch Setting up the Switch Internet Protocol over InfiniBand (IPoIB) Page 5 Configuration Configurations up to 24 End Nodes Configurations with More Than 24 End Nodes Page 6 Administration and Management Using Itutil Summary of Itutil Command Options Itutil Syntax Itutil Administrative Commands Itutil Management Commands Description: Without any options set, the command displays the basic details of all HP Fabric-configured host interfaces Syntax: itutil Restricting Options to One Interface Syntax: itutil -i [interface_name] Syntax Example: itutil -iib0 Description: This parameter only displays Syntax: itutil -t [transport_type] Syntax Example: itutil -tIB Displaying Statistics Syntax: itutil -s [-iinterface name|-ttransport type] Page Displaying Connectivity Information Using Itutil NOTE: This is a partial view of the itutil -ccommandoutput Displaying the List of End-Nodes Syntax: itutil -T [-iinterface name|-ttransport type] Example: itutil -T Viewing the Itutil Manpage Using NetTL Syntax: nettl -l [log class] -e[subsystem] [log class] Syntax Example: nettl -li w e -eIPOIB Stopping Logging and Tracing Syntax: nettl -sp IPoIB Administration and Management nwmgr Utility ifconfig Command lanadmin Commands Syntax: lanadmin -xall_if_info [any_ipoib_ppa] Syntax Example: lanadmin -xall_if_info Displaying IPoIB Interface MAC Address Syntax: lanadmin -a [ipoib_ppa] Syntax Example: lanadmin -a9000 lanadmin Command Help •Brief Help •Extended Help Brief Help Syntax: lanadmin -x -h [ipoib_ppa] Serviceguard and IPoIB Introduction Configuring IPoIB Interfaces for Serviceguard Clusters Identify all IPoIB ports that are configured for the application on each node Configure IP addresses for these interfaces using /etc/rc.config.d/netconf Verify the Switch Administration and Management Host Administration and Management” (page 56) This section addresses the following topics: •“CLI Overview” •“Using the CLI” CLI Overview Using the CLI Starting A CLI Session Entering CLI Modes Exiting CLI Modes Using Command Completion Command Abbreviation Quick Help Command-LineEditing Exiting the CLI Session Specifying the Card/Port Page Advanced Switch Setup Configuring the System Hostname Testing Network Connectivity Setting User Levels and Passwords Logging Onto the System Managing the Switch Page Configuration, Image, and Log Files Configuration, Image, and Log File Overview Upgrading Image Files •Out-of-Bandconnection •In-Bandconnection 2. Check the Image Version Refer to “Deleting Images Through the CLI ” (page 90) 1.Enter the privileged-executemode 3. Copy/Download the Image There are two ways that the software can be downloaded from the CLI: •Through an FTP server ((page 87)) •Through a TFTP server ((page 87)) Download From an FTP Server The “remote system” must be the IP address of the TFTP server. Syntax: HP-IB#copy tftp://<remote-system>/<filepath><filesystem>:<filename 3.(Optional) View the images by using the dir image command 4.Install a New Image To install an image file: •In the privileged-executemode, enter install image:file HP-IB#install image:image-file-name The name of the image file will change after it is installed oper-status boot-config 6. Reboot the System 1.Once the system is configured to use the new image, reboot the system yes copy “Reboot the System Through the CLI” (page 107) File Management Listing Configuration, Image, and Log Files Viewing Configuration Files Viewing Log Files Saving Configuration Files Saving for System Reboot Saving the Backup Configuration Specifying the Configuration to Use at System Reboot Saving and Copying Files Downloading Files to the System Deleting Configuration, Image, and Log Files Deleting Configuration Files To remove a configuration file: 1.Log in as the unrestricted user 3.Determine the configuration files in memory delete config: file Page Managing Log Files Understanding the Log Format Uploading Log Files Administering the System Switching User Identity Notifying Users Creating User Accounts Enter the username command and the 5.Repeat step 4 to change additional usernames and passwords 6.Exit the global-configurationmode 7.Use the show user command to verify changes Displaying User Information 4.Create the user account and user password The user name and password are alphanumeric strings up to 34 characters each ib-admin ib-passwd 5.Assign an SNMP community string to the new user account username HP-IB(config)#username ib-adminno User Account Configuration Commands Use the following commands and keywords to administer User Accounts: Table 6-7User Account Administrative Commands Using DNS Services Setting Administrative Roles Setting the System Clock Setting the NTP Servers to Maintain the System Clock Rebooting the System Page Page 7 Monitoring and Troubleshooting LED Color, Behavior, and Meaning HP Fabric Switch Cluster Connection LEDs HP Fabric Switch System Status Table 7-2Front System Status LED Indicator Table 7-3Rear System Status LED Indicator Power Supply Troubleshooting Ethernet Management Port LEDs HP Fabric HCA Connection LEDs Figure 7-3HP Fabric HCA Interface LEDs physical link status logical link status HCA LED location HCA Connection LED Colors and Meanings Determining Whether the HCA or Cable is Faulty Determining if the Switch is Faulty Next Steps HP-UXHost Troubleshooting Procedure Syntax example: itutil -T 5.Check communication path between nodes Syntax example: itutil -lfe80::2:c901:8a0:8a11 ib0 Syntax example: 6.Check adapter transport and performance statistics 7.Check connection details Syntax example: itutil -cfe80::2:c901:8a0:8a11 ib0 8.Check nettl logs Sample nettl Log Messages Description: Port failure Description: Firmware download failure Description: Monitoring and Troubleshooting IPoIB nettl Sample Log Output Sample IPoIB Logs Viewing Trace Packets in Raw Format Syntax: netfmt [file_name] Syntax Example: netfmt ipoib.TRC000 Viewing Trace Packets in Nice Format Linkloop Command Known Problems Health Monitoring About Logging and Tracing About HP Fabric Events About Tracing Troubleshooting the HP Fabric Network Ping Setting Trace Levels Replacing Individual Components Replacing a Power Supply Module A Specifications HCA Physical Specifications (AB286C) HCA Environmental Specifications (AB286C) HP Fabric HCA Electrical Specifications (AB286C) Page B Switch Command Line Interface Synopsis Syntax Command Modes Privilege Level Examples Defaults Related Commands show arp ethernet show arp IB show authentication show backplane show boot-config show card Table B-5show card Command Field Descriptions (continued) To display the configuration and status information for the card: Defaults: “show card” defaults to “show card all” Related Commands: “boot-config”(page 190) “card” (page 192) show card-inventory show clock show config show fan show host show ib show ib sm configuration subnet prefix | ]} The arguments associated with this command are described below: Table B-10 show ib Command Syntax Descriptions Command Modes: show ib sm multicast show ib sm neighbor show ib sm node subnet-prefix Table B-14show ib sm node subnet-prefixCommand Syntax Descriptions (continued) The fields in the show ib sm node output are described in the table below Table B-15show ib sm node subnet-prefixCommand Field Descriptions show ib sm partition show ib sm port Table B-17show ib sm port Command Field Descriptions (continued) Page HP-IB#show ib sm port subnet-prefixfe:80:00:00:00:00:00:00 Subnet Management Ports mkey : 00:00:00:00:00:00:00:00 gid-prefix :fe:80:00:00:00:00:00:00 lid : 14 master-sm-lid :1 cap-mask :00:10:02:48 diag-code :00:00 show ib sm service show ib sm switch Table B-20show ib sm switch Command Field Descriptions (continued) Subnet Management Switches subnet-prefix :fe:80:00:00:00:00:00:00 node-guid :00:05:ad:00:00:00:13:81 linear-fdb-cap :49152 random-fdb-cap :0 mcast-fdb-cap :1024 linear-fdb-top :1024 show ib-agent channel-adapter show ib-agentsummary show ib-agentswitch show ib-agentswitch linear-frd-info show ib-agentswitch all mcast-infolid show ib-agentswitch all node-info Table B-27show ib-agentswitch all node-infoCommand Field Descriptions HP-IB#show ib-agentswitch all node-info guid : 00:05:ad:00:00:00:13:7f type : switch lid : port-guid :00:05:ad:00:00:00:13:7f partition-cap :1 show ib-agentswitch all pkey-info show ib-agentswitch port-info show ib-agentswitch sl-vl-map show ib-agentswitch switch-info show interface ib Table B-31show interface ib Output Descriptions (continued) HP-IB#show interface ib 15/1 InfiniBand Interface Info port : 15/1 name : 15/1 show interface ib sm show interface ib sm statistics show interface mgmt-ethernet show interface mgmt-ib show interface mgmt-serial show ip show location show logging show ntp show power-supply show running-status show sensor show snmp show system-services show terminal show trace show user show version arp ib clear ib arp-cache default ip ip-over-ib-mtu ip domain-name ip ip-over-ib-mtu ip name-server-one ip name-server-two ip route ib sm subnet-prefix ib sm key no ib sm The syntax and arguments are described in the table below: Table B-46ib sm subnet-prefixSyntax Description Regarding Partitions: NOTE: HP recommends that the default partition be used for the cluster Table B-47Acceptable P_Key Values The following example removes a specified subnet manager The following example creates a Partition, and adds ib-agent channel-adapter ib-agentswitch interface ib link-trap name Administrative Commands This chapter lists and describes Administrative commands in alphabetical order Table B-54Administrative Commands Table B-54Administrative Commands (continued) action boot-config broadcast card clock configure copy delete dir disable enable exec exit ftp-serverenable gateway help history hostname install interface interface mgmt-ethernet interface mgmt-ib location login logging logout more ntp ping radius-server reload snmp-server telnet terminal length terminal time-out trace type username who write Page C How to Use Windows HyperTerminal HyperTerminal Prerequisite Loading HyperTerminal Configuring the HyperTerminal Connection Figure C-3Connect To Dialog Box Verifying the HyperTerminal Connection Figure C-5HyperTerminal Login Page Glossary IETF http://www.ietf.org/overview.html IOC I/O interface of the TCA is implemented by an I/O Controller datagrams across networks Multiple subnet managers per subnet are supported SMP Subnet Management Packet (SMP) on the fabric SNMP Target Channel Page Index Page Page Page Page Page (see also ports) ports) radius-servercommand, 189 RARP (Reverse Address Resolution Protocol), definition RDMA (Remote Direct Memory Access), 19, 224 overview show ib-agentswitch linear-frd-info show interface ib command, 132, 163–167 show interface ib sm command, 165–166 show interface ib sm statistics command, 166–167 show interface ib sma show interface mgmt-ethernet show interface mgmt-ethernetcommand, 167 show interface mgmt-ib