###################################################################### DELL OPENMANAGE(TM) SERVER ADMINISTRATOR 1.2.2 RELEASE NOTES ###################################################################### Dell OpenManage Server Administrator is a secure Web tool for managing individual Dell(TM) PowerEdge(TM) systems. Using this tool, an administrator with an Internet/intranet connection can manage a server from any location using either a graphical user interface (GUI) that runs in a Web browser or a command line interface (CLI) that is accessed through the command prompt. Server Administrator allows an administrator to do more than obtain status information. The administrator can also run diagnostics, configure BIOS and system settings, and perform system updates, which are management tasks that have traditionally been possible only from a central console or at the server itself. Integrated security leverages the server's operating system security infrastructure and uses the Secure Sockets Layer (SSL) encryption protocol. This file contains updated information for your "Dell OpenManage Server Administrator User's Guide" and any other technical documentation included with Server Administrator. NOTE: See the Dell OpenManage Array Manager "readme" file (located in the "am3xx" directory on the "Systems Management" CD) for supplemental information regarding the Storage Management Service. The online Server Administrator documentation includes the "User's Guide", the "Messages Reference Guide", the "CIM Reference Guide", the "Command Line Interface (CLI) Reference Guide", the "SNMP Reference Guide", and the "Compatibility Guide". You can access all the documentation from the product documentation CD or from the Dell Support website at "support.dell.com". See the Dell Support website ("support.dell.com") for the most current information. This document is organized into the following sections: * Server Administrator User Notes * Known issues for Server Administrator ###################################################################### WHAT'S NEW IN THIS RELEASE OF SERVER ADMINISTRATOR? ###################################################################### The primary purpose of Server Administrator 1.2.2, released on "Systems Management CD" 3.2.2 is to offer support for the new Dell PowerEdge 1600SC. A few minor features have been added to the Server Administrator Diagnostic Service. The PERC 4/im has now been added to the list of RAID controllers that can be tested on managed systems running supported Red Hat Linux or Microsoft(R) Windows(R) operating systems. Under supported Windows operating systems only, the Server Administrator Diagnostic Service can now also test hard drives when they are attached to the PERC 4/im RAID controller. In adding support for the Dell PowerEdge 1600SC, the Remote Access Service adds support for the DRAC III/XT remote access controller. A clarification has also been added to the readme for this release. Under "NOTES ABOUT THE UPDATE SERVICE" the first paragraph clearly states that BIOS updates are available for Server Administrator Windows-supported operating systems only. ###################################################################### "SYSTEMS MANAGEMENT" CD VERSION 3.2.2 SUPPORTED SYSTEMS ###################################################################### Dell OpenManage Server Administrator 1.2.2 is released on "Systems Management" CD version 3.2.2. The following PowerEdge systems are supported for "Systems Management" CD version 3.2.2: 300, 350, 500SC, 600SC, 1300, 1400, 1500SC, 1550, 1600SC, 1650, 2300, 2400, 2450, 2500, 2550, 2600, 2650, 4300, 4350, 4400, 4600, 6300, 6350, 6400, 6450, 6600, 6650, and 8450. ###################################################################### SERVER ADMINISTRATOR USER NOTES ###################################################################### This section provides information to help enhance your experience with Server Administrator in particular implementations and environments. ====================================================================== NOTES ABOUT INSTALLATION AND UNINSTALLATION ====================================================================== * You can find general installation information in the "Dell OpenManage Software Quick Installation Guide". You can find detailed installation information, including information about unattended installation, in the "Dell OpenManage Server Administrator User's Guide". Both of these documents are located on the product documentation CD. * Microsoft Internet Explorer Version 5.5 is available in the "\support\Windows\IE5.5" directory on the "Dell OpenManage Systems Management" CD. * Red Hat Linux RAID management tools for the PERC 4/Di controller are available in the "\support\Linux\RAID\lsi" directory on the "Systems Management" CD. These utilities support Red Hat Linux 7.1 SBE, 7.2, and 7.3. * Before you install Server Administrator, it is recommended that you uninstall any previous installations of Resolution Assistant. * You cannot deselect specific services during the custom installation and then install older versions of other service components such as the Instrumentation Service, Storage Management Service, or the Diagnostic Service. * For PowerEdge 8450 systems, allow the system to run for approximately 30 minutes before upgrading from a previous Dell OpenManage product release to this current release. * NetWare(R) command line interface utilities for the Dell PERC 2, PERC 2/Si, PERC 3/Si, and PERC 3/Di controllers are available in the "\support\Netware\RAID\adaptec" directory on the "Systems Management" CD. These utilities support both NetWare 5.1 and NetWare 6.0. * NetWare application utilities for the PERC 3/QC, PERC 3/DC, PERC 3/DCL, and PERC 3/SC controllers are available in the "\support\NetWare\RAID\ami" directory on the "Systems Management" CD. These utilities support both NetWare 5.1 and NetWare 6.0. * A Red Hat Linux management application for the PERC 2, PERC 2/Si, PERC 3/Si, and PERC 3/Di controllers is available in the "\support\Linux\RAID\adaptec" directory on the "Systems Management" CD. These utilities support both Red Hat Linux 7.2 and Red Hat Linux 7.1 SBE. * Red Hat Linux RAID management tools for the PERC 3/QC, PERC 3/DC, PERC 3/DCL, and PERC 3/SC controllers are available in the "\support\Linux\RAID\ami" directory on the "Systems Management" CD. These utilities support both Red Hat Linux 7.2 and Red Hat Linux 7.1 SBE. * LSI Logic offers a CIM solution application for monitoring 1020/1030-based PCI SCSI controllers, which includes the Dell PERC 4/im Integrated Mirroring solution. This application is available for Red Hat Linux operating system versions 7.1, 7.2, and 7.3 as supported by the platform. The application and installation instructions are located on the "Systems Management" CD in the "\ihv\cim_linux" directory. * The Managed System portion of Array Manager replaces the Windows Disk Management Service. Windows' Disk Management Service provides reporting and configuration of storage devices. To view "Disk Management" information in the Server Administrator GUI, install the Server Administrator's Storage Management Service. Installing the managed system Storage Management Service will automatically also install the management station component of this service. This will allow for the local management of the system hard drives. * The Storage Management Service requires that Array Manager version 3.1.1 or later is installed on the system to be managed as part of the Express Setup or Custom Setup managed system software installation from the "Systems Management" CD. When performing the managed system installation setup from the "Systems Management" CD, ensure that the "Storage Management" managed system software is selected for installation on the "Managed System Installation" window. ====================================================================== NOTES ABOUT THE INSTRUMENTATION SERVICE ====================================================================== * Redundant Memory: Redundant memory is a new feature supported on certain PowerEdge systems. It enables the use of reserve memory in the event of significant memory failure within the main memory modules. Server Administrator supports two types of redundant memory: spare bank and mirroring. Spare bank is a redundant memory implementation in which the contents of a memory module that is experiencing sufficient failures is copied into a spare module of the same type and capacity. The system then uses this spare bank module until the next reboot. The operating system does not discover or report the memory capacity allocated to the spare bank. Spare bank is supported on the following PowerEdge systems: 2600, 2650, 4600, 6600, and 6650. Mirroring is a redundant memory implementation in which the contents of the main memory modules are continually replicated in a duplicate set of memory modules of the same type and capacity. If sufficient failures occur in main memory, the system switches to using the mirrored memory. The system uses the mirrored memory modules until the next reboot. The operating system does not discover or report half of the installed system memory in this implementation. Mirroring is supported on the following PowerEdge systems: 6600 and 6650. * Virtual Memory The virtual memory value for systems running Microsoft(R) Windows NT(R) and Windows(R) 2000 always shows 2064404. This value is the maximum virtual memory available to any current process. * Users can set minimum and maximum temperature probe threshold settings for an external chassis to whole numbers only. If users attempt to set either the minimum or maximum temperature probe threshold setting to a number that contains a decimal, only the whole number before the decimal place is saved as the threshold setting. * Setting the boot sequence to "Device List" on the BIOS "Setup" tab in Server Administrator results in the following boot sequence: diskette, IDE CD drive, hard drive, option ROMs (if the devices are available). * CMOS Clear Operations for PowerEdge 2450 Systems: For PowerEdge 2450 systems that are new or have undergone a CMOS clear operation, the Primary SCSI setting is reported as "unknown" or "off" even after setting the onboard SCSI controller to "ON" in the BIOS setup. To ensure that value is reported correctly, click "Primary SCSI", set the state appropriately, and then click "Apply". * Memory Error Counts: The memory error counts have been split into "Single-Bit Correction Count" and "Multibit Error Count". Clearing the memory errors is no longer applicable as a user function and is no longer supported. * Dell MIB Files for PowerEdge Systems: Dell MIB files for PowerEdge systems allow customers to obtain and verify information provided by supported software agents. The current MIB files supported by PowerEdge software agents are located at "\support\dcpemibs" on the "Systems Management" CD. NOTE: A MIB-II-compliant, SNMP-supported network management station is required to compile and browse the MIB files. * SNMP Agent Configuration: To manage a system running Server Administrator using SNMP management applications such as IT Assistant, you must configure the SNMP agent on the Server Administrator system properly. For IT Assistant to retrieve management information from a system running Server Administrator, the "Get" community name used by IT Assistant must match a community name that allows "read" access on the system running Server Administrator. For IT Assistant to modify information or perform actions on a system running Server Administrator, the "Set" community name used by IT Assistant must match a community name that allows "write" access on the system running Server Administrator. For IT Assistant to receive traps (asynchronous event notifications) from a system running Server Administrator, the system running Server Administrator must be configured to send traps to the system running IT Assistant. Failure to configure these features properly may result in the inability to manage your system using SNMP or to receive important event notifications from your system. Review the SNMP configuration information contained within the "Dell OpenManage Server Administrator User's Guide" and information included with your operating system. * Dynamic Kernel Support on Red Hat Linux: Table 1. Precompiled Kernels Supported by Server Administrator With Prebuilt Device Drivers ----------------------------------------------------- | Operating System | Kernel | ----------------------------------------------------- | Red Hat Linux 7.2 | 2.4.7-10 | | | 2.4.7-10smp | | | 2.4.7-10enterprise | ----------------------------------------------------- | Red Hat Linux | 2.4.9-e.3 | | Advanced Server 2.1 | 2.4.9-e.3smp | | | 2.4.9-e.3enterprise | ----------------------------------------------------- | Red Hat Linux 7.3 | 2.4.18-4 | | | 2.4.18-4smp | | | 2.4.18-4bigmem | ----------------------------------------------------- The Instrumentation Service provides prebuilt device drivers for the precompiled kernels listed above in Table 1. When the Instrumentation Service detects that none of its prebuilt device drivers supports the running kernel, it attempts to build a device driver for that kernel. This feature of Server Administrator is called dynamic kernel support. Server Administrator's dynamic kernel support builds a device driver when needed during the Instrumentation Service installation or startup. If the running kernel is not one of those listed in Table 1 or has been reconfigured and recompiled in such a way that none of the prebuilt device drivers support the kernel, dynamic kernel support must be used to support the running kernel. For dynamic kernel support to be used, certain prerequisites must be met. See the "Dell OpenManage Server Administrator User's Guide" for more information about dynamic kernel support. * The "Dell OpenManage Server Administrator User's Guide" lists the "kernel-headers" RPM package as a Dynamic Kernel Support prerequisite. On Red Hat Linux 7.3, the "glibc-kernheaders" RPM package replaces the "kernel-headers" RPM package. * SNMP Access Control Configuration on Red Hat Linux: The MIB branch implemented by the Instrumentation Service is identified by the OID "1.3.6.1.4.1.674.10892.1". Management applications must have access to this branch of the MIB tree to manage systems running the Instrumentation Service. The default SNMP configuration installed by the ucd-snmp package provides read-only access only to the MIB-II "system" branch (identified by the OID 1.3.6.1.2.1.1) of the MIB tree for the "public" community. This configuration does not allow management applications to retrieve or change Instrumentation Service information. If the Instrumentation Service detects this configuration during installation, it attempts to modify the SNMP configuration to provide read-only access to the entire MIB tree for the "public" community. * Firewall Configuration on Red Hat Linux: If you select High or Medium firewall security when installing Red Hat Linux, the SNMP port on all external network interfaces is closed by default. To enable SNMP management applications such as IT Assistant to discover and retrieve information from the Instrumentation Service, the SNMP port on at least one external network interface must be open. If the Instrumentation Service detects that the SNMP port is not open in the firewall for any external network interface, the Instrumentation Service displays a warning message and logs a message to the system log. You can open the SNMP port by disabling the firewall, opening an entire external network interface in the firewall, or opening the SNMP port for at least one external network interface in the firewall. You can perform this action before or after the Instrumentation Service is started. To open the SNMP port using one of the previously described methods, perform the following steps: 1. At the Red Hat Linux command prompt, type "setup" and press to start the Text Mode Setup Utility. The "Choose a Tool" menu appears. 2. Select "Firewall Configuration" using the down arrow and press . The "Firewall Configuration" screen appears. 3. Select the Security Level by tabbing to it and pressing the spacebar. The selected Security Level is indicated by an asterisk. NOTE: Press for more information about the firewall security levels. The default SNMP port number is 161. a. To disable the firewall, select "No firewall" and go to step 7. b. To open an entire network interface or the SNMP port, select "High" or "Medium" and continue with step 4. 4. Tab to "Customize" and press . The "Firewall Configuration - Customize" screen appears. 5. Select whether to open an entire network interface or just the SNMP port on all network interfaces. a. To open an entire network interface, tab to one of the Trusted Devices and press the spacebar. An asterisk in the box to the left of the device name indicates that the entire interface will be opened. b. To open just the SNMP port on all network interfaces, tab to "Other ports" and type "snmp:udp". 6. Tab to "OK" and press . The "Firewall Configuration" screen appears. 7. Tab to "OK" and press . The "Choose a Tool" menu appears. 8. Tab to "Quit" and press . * Remote Flash BIOS Update - Red Hat Linux If you plan to perform a remote flash BIOS update on a PowerEdge 1500SC, 1550, 1650, 2500, or 2550 system that is running Red Hat Linux with a kernel that includes Symmetrical Multi-Processing (SMP) support, you must perform one of the following series of steps for the update to occur properly. - If you are using the Linux Loader (LILO), perform the following steps: 1. Edit the LILO configuration file "/etc/lilo.conf". 2. Add the following line after the "image" line for any kernels that include SMP support: append="reboot=bios" For example, suppose the following line is in the configuration file: image=/boot/vmlinuz-2.4.3-6smp You would add the second line indented below the first: image=/boot/vmlinuz-2.4.3-6smp append="reboot=bios" 3. After updating the LILO configuration file, run the following command for the change to take effect: lilo 4. Reboot the system. 5. Perform the remote flash BIOS update. - If you are using the GNU Grand Unified Bootloader (GRUB), perform the following steps: 1. Edit the GRUB configuration file "/etc/grub.conf". 2. Append "reboot=bios" to the "kernel" line for any kernels that include SMP support. For example, suppose the following line is in the configuration file: kernel /vmlinuz-2.4.7-10enterprise ro root=/dev/sda3 You would append "reboot=bios" to the end: kernel /vmlinuz-2.4.7-10enterprise ro root=/dev/sda3 reboot=bios 3. Reboot the system. 4. Perform the remote flash BIOS update. * Alert Actions - Red Hat Linux - When you set Alert Actions for an event, you can specify the action to "display an alert on the server." To perform that action, Server Administrator writes a message to the console. If the Server Administrator system is running X Windows, you will not see that message by default. To see the alert message when X Windows is running, start "xconsole" before the event occurs. - When you set Alert Actions for an event, you can specify the action to "broadcast a message". To perform that action, Server Administrator executes the "wall" command, which sends the message to everybody logged in with their message permission set to "yes". If the Server Administrator system is running X Windows, you will not see that message by default. To see the broadcast message when X Windows is running, start a terminal such as "gnome-terminal" before the event occurs. -When you set Alert Actions for an event, you can specify the action to "execute an application". There are limitations on the applications that Server Administrator can execute. Follow these guidelines to ensure proper execution: 1) Do not specify X Windows-based applications because Server Administrator cannot execute such applications properly. 2) Do not specify applications that require input from the user because Server Administrator cannot execute such applications properly. 3) Redirect "stdout" and "stderr" to a file when specifying the application so that you can see any output or error messages. 4) If you want to execute multiple applications (or commands) for an alert, create a script to do that and put the full path to the script in the "application to execute" box. Example 1: ps -ef >/tmp/psout.txt 2>&1 The command in Example 1 executes the application "ps", redirects "stdout" to the file "/tmp/psout.txt", and redirects "stderr" to the same file as "stdout". Example 2: mail -s "Server Alert" admin /tmp/mailout.txt 2>&1 The command in Example 2 executes the application "mail" to send the message contained in the file "/tmp/alertmsg.txt" to Linux user "admin" with the subject "Server Alert". The file "/tmp/alertmsg.txt" must be created by the user before the event occurs. "stdout" and "stderr" are redirected to the file "/tmp/mailout.txt" in case an error occurs. ====================================================================== NOTES ABOUT THE UPDATE SERVICE ====================================================================== * BIOS updates are currently available only on Microsoft Windows operating systems supported by Server Administrator. Linux and Novell(R) NetWare operating systems do not currently support BIOS updates from Server Administrator. * BIOS updates are not supported for PowerEdge 350 and 8450 systems. * Firmware updates are not supported for PowerEdge 1500SC, 1550, and 8450 systems. NOTES: PowerEdge 2550 system BIOS updates are supported if the BIOS revision is later than or equal to A06. PowerEdge 500SC system BIOS updates are supported if the BIOS revision is later than or equal to A05. PowerEdge 1650 system BIOS updates are supported if the BIOS revision is later than or equal to A05. The "Last Update Date" field indicates the date that the BIOS was flashed through Server Administrator or other Dell OpenManage products. This date does not reflect the last BIOS update that occurred using a diskette. * The Update Service will not perform PERC, CERC, or ROMB firmware and device driver updates unless the Array Manager agent is installed. The Array Manager agent is installed with Server Administrator by default. * The minimum required level of PERC device driver for Windows NT is 4.1.1.0. * The minimum required level of PERC device driver for Windows 2000 is 5.2.2.0 * The minimum required level of firmware for PERC 3/DC/DCL/QC is 1.61j. * The minimum required level of firmware for PERC 3/SC is 1.63. * The Update Service 1.2 supports only the PERC 3/DC/DCL/QC and PERC 3/SC. * The Update Service 1.3 supports the PERC 4/Di on the PowerEdge 2600 only. * If you receive an error message after updating your firmware that states, "Could not start the Dell OpenManage Server Agent Service [...]", you will need to reboot your system. * To perform a PERC 4/im firmware update in the Server Administrator GUI, select "System" in the system tree. Click the "Update" tab. The "Version Report" appears. For the BIOS component, select "Perform Update" and supply the required update package information. To perform a PERC 4/im firmware update in the Server Administrator command line interface, use the "omupdate biosupdate path=" command. ====================================================================== NOTES ABOUT THE DIAGNOSTIC SERVICE ====================================================================== * Server Administrator diagnostics specifically support the following Dell-supplied hardware devices: - CD drives - DVD drives - CD-RW/DVD drives - Diskette and LS drives - Network interface cards - Modems - The following tape drives, autoloaders, and libraries: (All tape drives attached via a PowerVault(TM) tape rack enclosure.) For Windows: - Tape drives: PV 100T IDE tape drives, PV 100T DDS4, PV 110T DLT7000, PV 110T DLT1, PV 110T VS80, PV 110T SDLT, PV 110T LTO, all drives in supported autoloaders and libraries (see below) - Tape autoloaders and libraries: PV 120T DDS4, PV 120T DLT7000, PV 120T DLT1, PV 122T VS80, PV 128T LTO, PV 128T SDLT, PV 132T LTO, PV 136T LTO, PV 136T SDLT (In order to properly run diagnostics on tape drives and automation devices, Dell-qualified device drivers should be installed. See the hardware documentation or "support.dell.com" for more information on device drivers.) For Red Hat Linux: - Tests are supported for all tape drives. - Tape autoloaders and libraries are not supported. - Video controllers on systems running the Windows 2000 and Red Hat Linux operating systems - SCSI controllers - The following RAID controllers: For Windows: PERC 2/DC, PERC 2, PERC 2/Si, PERC 3/Si, PERC 3/Di, PERC 3/QC, PERC 3/DC, PERC 3/DCL, PERC 3/SC, CERC ATA 100, PERC 4/Di, and PERC 4/im (See firmware requirements in the Compatibility Guide.) For Red Hat Linux: PERC 2/DC, PERC 2, PERC 2/Si, PERC 3/Si, PERC 3/Di, PERC 3/QC, PERC 3/DC, PERC 3/DCL, PERC 3/SC, PERC 4/Di, and PERC 4/im (See firmware requirements in the Compatibility Guide.) - Hard drives attached to the following RAID controllers (Windows only): PERC 2/DC, PERC 2, PERC 2/Si, PERC 3/Si, PERC 3/Di, PERC 3/QC, PERC 3/DC, PERC 3/DCL, PERC 3/SC, CERC ATA 100, PERC 4/Di, and PERC 4/im (See firmware requirements in the Compatibility Guide.) - All hard drives attached to RAID controllers (Red Hat Linux) * When upgrading from Server Diagnostics 1.1 to the Diagnostic Service, the history file is saved in the "\bak" directory located in the root installation directory. The filename of this history log is "result.xml". * If you are using Server Administrator on a system that is completely isolated on a network (cannot ping any other systems) you may experience performance problems and may receive inconsistent results from the Diagnostics Service if 1) the "Automatically detect settings" checkbox is on in the Local Area Network proxy settings page of Internet Explorer, and 2) your version of Internet Explorer is earlier than 5.5. To correct this issue, uncheck the "Automatically detect settings" check box. ====================================================================== NOTES ABOUT THE REMOTE ACCESS SERVICE ====================================================================== This feature is supported on the PowerEdge 1600SC, 1650, 2600, 2650, 4600, 6600, and 6650 systems only. It enables remote access to a server that has lost its network connection or that has become unresponsive. In this release of Server Administrator, the Remote Access Service uses one of the following remote access controllers (RACs): the Dell Remote Access Card (DRAC) III, the DRAC III/XT, the embedded remote access (ERA), or the embedded remote access option (ERA/O). RACs also have their own command line interface that is accessed through the "racadm" command. You can add "racadm" commands to a batch or script file to automate various user tasks. To limit the stress load on the managed system and RAC, you should add "sleep" or "delay" commands of one or two seconds between the individual "racadm" commands. For systems running supported Red Hat Linux operating systems, the Remote Access Service does not support Netscape Navigator. ###################################################################### KNOWN ISSUES FOR SERVER ADMINISTRATOR ###################################################################### This section provides information on open issues with this release of Server Administrator. ====================================================================== ISSUES FOR SERVER ADMINISTRATOR RUNNING ON ALL SUPPORTED OPERATING SYSTEMS. ====================================================================== * Clicking the browser "Back" button may not display the correct page with respect to the Server Administrator component tree, tabs, or tab menus. Example: 1. Click "Main System Chassis" under "System Component Tree". 2. Click "Fans" under "Main System Chassis". 3. Click "Alert Actions" on the tab menu bar. 4. Click the "Back" button on the browser. This action takes you to the "Fan Probes" page, leaving the tab menu bar unchanged. This action places the "Fan Probes" page under "Alert Actions" on the tab menu bar instead of "Fan Probes" on the tab menu bar. * Selecting the boot sequence under the BIOS "Setup" tab will not reenable boot devices that have been previously disabled in the System Setup Program. * The links on the Server Administrator home page might lock up after repeated random clicking. To resolve this situation, refresh the browser by pressing or by clicking the browser refresh button. * All unsecure HTTP requests to Server Administrator receive an invalid response. Server Administrator runs only one instance of the Web server, which is secure. Make all connections through https://:. Any http://: request for connection with the server receives an invalid response. * If the browser used with Server Administrator indicates that it cannot display a page or perform an action, ensure that the browser is in online mode. To go online, perform one of the following steps: - If you are using Netscape Navigator, click "File" on the menu bar, point to "Offline", and select "Work Online". OR - If you are using Internet Explorer, click "File" on the menu bar and deselect the "Work Offline" option. When "Work Offline" is selected, a check displays to the left of the option on the "File" menu. * If you are using Netscape Navigator and you receive the error message, "Netscape was unable to connect to the network because you are offline...", Server Administrator is in offline mode. Go online to avoid this message. * If Internet Explorer prompts you to "Work Offline", "Connect", or "Try Again", always select "Connect" or "Try Again". Do not select "Work Offline". * For PowerEdge 8450 systems running Array Manager, allow the system to run for approximately 30 minutes before logging in to the Server Administrator Web server. If you do not wait 30 minutes, Array Manager might not display on the Server Administrator home page or Server Administrator might hang. * When setting dates in the "Asset Information" section of the Server Administrator home page, the current time will be appended to the date. When setting dates with the Command Line Interface, the appended time will be noon. * If Network Adapter Teaming is installed and enabled on your system, Server Administrator will not display the IP address or other connection-related data for the individual network adapters. The connection status and IP address belong to the virtual adapter created by the Teaming software. The virtual adapter is not currently displayed in Server Administrator. * For systems running a supported Red Hat Linux operating system, kernel driver messages such as "AAC_ChardevOpen" or "Attached scs CD-ROM sr0 at scsi0" are sometimes displayed in the console at the login prompt. These messages display in the console when the driver initialization is delayed by the installation of Server Administrator services and can be safely ignored. ====================================================================== ISSUES FOR SERVER ADMINISTRATOR RUNNING ON ALL NOVELL NETWARE OPERATING SYSTEMS ====================================================================== * When you log in to Server Administrator for the first time on a system running NetWare 5.1, the information displayed on the Server Administrator home page might be incomplete. If this occurs, press to refresh the browser window. Complete information for your managed system should now be visible on the Server Administrator home page. *If you shut down Server Administrator on a system running NetWare using the "sys:\system\dell\omanage\iws\bin\netware\omastop.ncf" command, the shutdown script will unload the Java Virtual Machine on the NetWare system. As a result, all applications using the java NLM will also be shut down. * Under "Network information", the IP address displays as "Not Obtained". ====================================================================== ISSUES FOR SERVER ADMINISTRATOR RUNNING ON THE NOVELL NETWARE 5.1 OPERATING SYSTEM ====================================================================== * If you receive an "NDSComponentException" message while starting Server Administrator, perform one of the following steps: - If the default certificate is corrupt or does not exist, see the "Web Server and Security Setup" section of the "User's Guide" to create a new certificate. - If the host name is not present in "client_properties.ini" and "server_properties.ini", perform the following steps: 1. Open the "client_properties.ini" file in "SYS:\system\dell\OpenManage\IWS\config". 2. Change the to your system name on the line "nssl.keystore = certificate name - hostname". 3. Open the "server_properties.ini" file in "SYS:\system\dell\OpenManage\IWS\config". 4. Change the and to your system's name on the line "nssl.keystore = certificate name - hostname". * If you receive an "SSLComponentException" message while starting the Server Administrator, perform the following steps (not setting "classpath" correctly to "SYS:\java\lib\nssl1.1_exp.jar" might be a reason for this exception): 1. Ensure that "nssl1.1_exp.jar" or "nssl1.1_dom.jar" exists in the "SYS:\java\lib" directory. 2. Set classpath to "SYS:\java\lib\nssl1.1_exp.jar" or "SYS:\java\lib\nssl1.1_dom.jar" by typing one of the following commands: envset classpath=$classpath;SYS:\java\lib\nssl1.1_exp.jar envset classpath=$classpath;SYS:\java\lib\nssl1.1_dom.jar NOTE: If the previous solutions do not match your configuration, see the Java SSL FAQ at "developer.novell.com" for solutions that might match your configuration. * When installing on Novell NetWare 5.1 with Java Virtual Machine (JVM) 1.1.7B and Support Pack 3, you might receive a java.io exception after you finish the installation and restart the managed system. To correct this issue, go to the Novell support website at "support.novell.com" and download the latest JVM 1.1.7B dated after 07/23/2001. TID number for this incident on the Novell website is 10066646. * If you are using the NOVONYX Web server in NetWare 5.1 with Server Administrator, then Server Administrator Web server may stop responding after some time. To avoid this behavior, ensure that the NOVONYX Web server starts after the Server Administrator Web server by performing the following steps: 1. Find the following line in the "AUTOEXEC.NCF" file: NSWEB This line starts the NOVONYX Web server. 2. Move this line to either the end of the "AUTOEXEC.NCF" file or to another location that is after (below) the following line: SYS:\SYSTEM\DELL\OMANAGE\IWS\BIN\NETWARE\OMASTART.NCF This line starts the Server Administrator Web server. * If you install Server Administrator on a system running NetWare 5.1 and Java 1.1.7B, the Server Administrator pages do not display correctly in Netscape 6.1 or 6.2. * If you install Server Administrator on a system running NetWare 5.1 and Java 1.1.7B, the Server Administrator service may hang if you load ConsoleOne on the NetWare system. You will have to run the following commands at the NetWare command prompt to restart the server after you have exited from ConsoleOne. SYS:\SYSTEM\DELL\OMANAGE\IWS\BIN\NETWARE\OMASTOP SYS:\SYSTEM\DELL\OMANAGE\IWS\BIN\NETWARE\OMASTART ====================================================================== ISSUES FOR SERVER ADMINISTRATOR RUNNING ON NOVELL NETWARE 6.0 OPERATING SYSTEMS ====================================================================== * This version of Server Administrator does not support Adaptec CI/O on systems running Novell NetWare 6.0. NOTE: Although the Adaptec CI/O option is available from the main installation menu for systems running NetWare 6.0, the installation program blocks this component. This menu item is intended to accommodate future versions of the software. ====================================================================== ISSUES FOR INSTALLATION AND UNINSTALLATION ====================================================================== The following are open issues and workarounds, where available, for Server Administrator on systems that are running Windows NT, Windows 2000, Red Hat Linux, or the Novell NetWare operating system. ---------------------------------------------------------------------- INSTALLATION AND UNINSTALLATION ISSUES FOR ALL SUPPORTED OPERATING SYSTEMS ---------------------------------------------------------------------- * Versions of Array Manager that were installed from the "Dell OpenManage Applications" CD version 1.6 must be uninstalled through the operating system (such as "Add/Remove Program" in Windows) before upgrading to the current version of Array Manager using the "Systems Management" CD version 3.1.1. * For systems running a supported Windows or Red Hat Linux operating system, some managed system components cannot be uninstalled using the Systems Management CD. These components must be uninstalled using the operating system ("Add/Remove Program" for Windows; running the appropriate RPM command for Red Hat Linux). * If you already have Adaptec Fast Console installed on your system, you must uninstall this application before installing the Storage Management service. * If you have Dell OpenManage Resolution Assistant installed on your system, you must manually uninstall it before installing any systems management applications. * If you upgrade from Server Agent 4.x to Server Administrator 1.0, the advanced installation settings for Server Administrator may not be available at the time of installation. This event occurs only when you upgrade the management station components first and then upgrade the managed system components. If the advanced settings are unavailable, you can set them when the installation finishes. ---------------------------------------------------------------------- INSTALLATION AND UNINSTALLATION ISSUES FOR SUPPORTED WINDOWS OPERATING SYSTEMS ---------------------------------------------------------------------- * On systems running Windows operating systems, when you uninstall all the management station components using the "Add/Remove Program" function, an empty folder named "Dell OpenManage Applications" might be left in the "Start" menu under "Programs". It is safe to manually remove this folder. * When performing an unattended uninstall of managed system components via the "setup.exe /u" command, it is strongly recommended that you reboot your system after the uninstall. Without the reboot, the Windows Disk Management tool may not function. * Before installing Citrix Metaframe (all versions), you must uninstall any previous installations of Server Administrator. Errors may exist in the registry after a Citrix Metaframe installation. ---------------------------------------------------------------------- INSTALLATION AND UNINSTALLATION ISSUES FOR SUPPORTED WINDOWS NT OPERATING SYSTEMS ---------------------------------------------------------------------- * Due to a Windows security policy issue, you can only uninstall Dell OpenManage Server Agent if you are logged in with the same user ID that was originally used when Server Agent was installed. To upgrade to Server Administrator from any version of Server Agent, you must log in to the system to be managed with the same user ID that was used when Server Agent was installed. If this user account is not available, you must reinstall your current version of Server Agent, and then upgrade to Server Administrator. * Uninstallation of some independent hardware vendor (IHV) utilities such as Adaptec CIO and LSI DMI Browser using "Add/Remove Program" may leave a few files behind. It is safe to manually delete these files. * The Windows installer service does not allow installation using Terminal Services when mapping to a remote CD drive. Because the Windows Installer runs as a system service, when an administrator installs from a remote session on Windows 2000, the installation must use a universal naming convention (UNC) path and not a mapped drive letter. In this case, drive letters mapped in a remote session may not be visible to the Windows Installer service. * If you install Intel(R) Client Instrumentation for DMI and SNMP from the "Dell OpenManage Systems Management" CD 3.0 on a system running Windows NT 4.0 and then upgrade the operating system to Windows 2000, you see the following error message in the event log: "Intel Client Instrumentation for DMI and SNMP hung on starting". Intel Client Instrumentation for DMI and SNMP is not supported on Windows 2000. * When you install the Adaptec CI/O Management package, the following warning appears in the Windows event log: "The SNMP Service is ignoring extension agent key SOFTWARE\...\SNMPAgent\CurrentVersion because it is missing or misconfigured". You can safely ignore this message. * During installation, the installation procedure may appear to hang at around 40 percent complete. The install module is waiting for the SNMP service to stop before proceeding. This waiting period can last up to 3 minutes. ---------------------------------------------------------------------- INSTALLATION AND UNINSTALLATION ISSUES FOR SUPPORTED RED HAT LINUX OPERATING SYSTEMS ---------------------------------------------------------------------- * Installing Server Administrator on systems running supported Red Hat Linux operating systems does not require that you reboot the system to complete installation. * When "/var" is full you can receive the error message, "Operating system support for SNMP is not installed....". This message may or may not be correct. Regardless, the instrumentation install will fail. Free space in "/var" is required for install to function properly. Free up space in "/var" and reinstall. ---------------------------------------------------------------------- INSTALLATION AND UNINSTALLATION ISSUES FOR SUPPORTED NOVELL NETWARE OPERATING SYSTEMS ---------------------------------------------------------------------- * Installing Server Administrator on systems running supported Novell NetWare operating systems requires that you reboot the system to complete the installation. * For NetWare systems running Adaptec CI/O: When the Adaptec CI/O Administrator is uninstalled from a system running NetWare 5.x, the following message appears on the System Console: "Unloading CI/O Management Service (IOMGR) is not recommended. Unload module anyway? n" When the Instrumentation Service uninstall program detects the Adaptec CI/O Administrator on your system, the program switches to the System Console to provide this message to the user. To unload the Adaptec CI/O Administrator and allow the uninstall program to continue, press . * For managed systems running NetWare 5.1 or 6.0: Before running the Server Administrator installation in silent mode, ensure that the NWconfig utility is not running. The Server Administrator installation depends on the NWconfig utility; if the NWconfig utility is already running when you start a silent installation, the NWconfig utility aborts and displays the message, "This module is already loaded and cannot be loaded more than once". ====================================================================== ISSUES FOR INSTRUMENTATION ====================================================================== The following are open issues and workarounds, where available, for Server Administrator on systems that are running Windows NT, Windows 2000, Red Hat Linux, or the Novell NetWare operating system. ---------------------------------------------------------------------- INSTRUMENTATION ISSUES FOR ALL OPERATING SYSTEMS ---------------------------------------------------------------------- * The NIC configuration information within the Server Administrator "BIOS Setup" screen may be inaccurate for embedded NICs. Using the Server Administrator "BIOS Setup" screen to enable or disable the NIC might produce unexpected results. It is recommended that you perform all NIC configurations through the actual System Setup screen that is available by pressing while a system is booting. This problem is present in the CLI interface and in IT Assistant. * If a system with spare bank memory enabled enters a "redundancy lost" state, it may not be apparent which DIMM is the cause. If you cannot determine which DIMM to replace, see the "switch to spare memory bank detected" log entry in the ESM system log to find which DIMM failed. * Setting any warning threshold to a default value is not supported on the PowerEdge 1650, 2600, 2650, 4600, 6600, and 6650 systems. Attempting to do so results in an error message. * Warning thresholds that are modified on the following systems are reset to their default values if the system is power cycled or loses AC power: the PowerEdge 1650, 2600, 2650, 4600, 6600, and 6650. If you change your default warning threshold values and notice that they have been reset to default values after a reboot, then you must add the relevant "omconfig" commands to your system startup sequence. For example, if you want to set the warning thresholds for fan zero to a minimum of 1000 and a maximum of 5300, add the following command to your startup sequence: omconfig chassis fans index=0 minwarnthresh=1000 maxwarnthresh=5300 Ensure that the "omconfig" command is invoked after system services have started. See the "Dell OpenManage Command Line Interface User's Guide" for more information. ---------------------------------------------------------------------- INSTRUMENTATION ISSUES FOR THE WINDOWS NT AND WINDOWS 2000 OPERATING SYSTEMS ---------------------------------------------------------------------- * To use the Instrumentation Service SNMP Subagent with HP OpenView Network Node Manager (NNM) Version 6.1, you must install the HP OpenView consolidated patch that contains the NNM_00496 patch on your system. To download the patch, go to the Hewlett Packard website located at "www.hp.com". * In some circumstances the Server Administrator CIM interface may not be properly registered with the operating system and will not function. To resolve this situation, perform the following steps: 1. Open a command prompt window. 2. Change to the installation directory, which is typically "C:\Program Files\Dell\OpenManage\OMSA\bin". 3. Run the command "regsvr32 dccim32.dll". ---------------------------------------------------------------------- INSTRUMENTATION ISSUES FOR THE RED HAT LINUX OPERATING SYSTEM ---------------------------------------------------------------------- *Server Administrator does not show the IP address and connection status for a Dell PowerEdge if the system: a) has a Broadcom Gigabit Ethernet card installed and b) is running the Red Hat Linux 7.3 operating system. To correct this problem, you must install the bcm5700 driver on your system. 1. Place the "Dell OpenManage Server Assistant" (DSA) CD in a system running a Windows operating system. 2. Locate the "Broadcom: NetXtreme Family of Adapters (Driver) 6.0" for Red Hat Linux 7.3 for your Dell PowerEdge system and copy the files to a diskette. NOTE: The "DSA" CD contains a precompiled version of the bcm5700 driver for the 2.4.18-4 Red Hat Linux kernel. The CD also contains the source files if you have updated your operating system to a later kernel. Use the ".src.rpm" for the NIC to compile the driver for your version of the kernel. Instructions for compiling the driver for a different kernel version are included with the driver source. 3. Using the directions that come with the driver, install the bcm5700 driver. 4. View the "/etc/modules.conf" file. If there are entries that refer to "tg3", substitute all occurrences of "tg3" with "bcm5700". For example, edit the following line: "alias eth1 tg3" to read: "alias eth1 bcm5700" 5. Edit the "modules.pcimap" file to replace all occurrences of "tg3" with the "bcm5700" string for the Broadcom NIC. * "Current Speed" and "Minimum Failure" are the only valid fan threshold values for storage systems attached to managed systems running the Red Hat Linux operating system. All other values are displayed as "unknown". * PowerVault 20xS, 21xS, and 22xS storage systems attached to managed systems running the Red Hat Linux operating system cannot be managed by Server Administrator using the following RAID controllers: - PERC 2 - PERC 2/Di - PERC 2/Si - PERC 3/Di - PERC 3/Si * The following information replaces the section "Copying a Dynamically Built Device Driver to Systems Running the Same Kernel" in the "Dell OpenManage Server Administrator User's Guide": Copying a Dynamically Built Device Driver to Systems Running the Same Kernel When Server Administrator dynamically builds a device driver for the running kernel, it installs the device driver into the directory "/lib/modules//misc", where is the kernel name returned by typing "uname -r". If you have a system running the same kernel for which a device driver was built, you can copy the newly built device driver to the directory "/var/omsa/dks/" on the other system for use by Server Administrator. This action allows Server Administrator to use its Dynamic Kernel Support feature on multiple systems without having to install the kernel source on every system. For example: System A is running a kernel that is not supported by one of the Server Administrator prebuilt device drivers. System B is running the same kernel. Perform the following steps to build a device driver on system A and copy the device driver to system B for use by Server Administrator: 1. Ensure that the Dynamic Kernel Support prerequisites are met on system A. 2. Install Server Administrator on system A. Server Administrator builds a device driver for the kernel running on system A during installation. 3. Type "uname -r" on system A to determine the name of the running kernel. 4. Copy the file "/lib/modules//misc/esm.o" from system A to the directory "/var/omsa/dks/" on system B, where is the kernel name returned by typing "uname -r" in step 3. NOTE: You might have to create the directory "/var/omsa/dks/" on System B. For example, if the kernel name is "1.2.3-4smp", you can create the directory by typing "mkdir -p /var/omsa/dks/1.2.3-4smp". 5. Install Server Administrator on system B. Server Administrator detects that the device driver you copied to the directory "/var/omsa/dks/" supports the running kernel and uses that device driver. NOTE: If this procedure is used to install Server Administrator, this procedure must be used to upgrade Server Administrator if the new version of Server Administrator does not support the running kernel with a prebuilt device driver. NOTE: When Server Administrator is uninstalled from System B, the process does not remove the "/var/omsa/dks//esm.o" file that you copied to system B. You must remove the file if it is no longer needed. ---------------------------------------------------------------------- INSTRUMENTATION ISSUES FOR THE NOVELL NETWARE OPERATING SYSTEM ---------------------------------------------------------------------- * The Instrumentation Service might incorrectly report NetWare version 5.1 as version 5.0. * On managed systems running the Novell NetWare operating system, the Instrumentation Service loads multiple NetWare Loadable Modules (NLMs) at system startup. Some NLMs will automatically uninstall because they are not required on a particular system. The NLM uninstall process occurs approximately two minutes after you boot the system. During this uninstall process, a module unloading message appears on the Novell NetWare system console. * After installing the Instrumentation Service, the user reboots the system to make the newly installed managed system services available for use. While rebooting, the system might hang if Adaptec CI/O management software is installed. To avoid a hung system during reboot, perform the following steps: 1. Remove "dcstart.ncf" from "autoexec.ncf". 2. After the system reboots, run "dcstart.ncf" from the NetWare command line. NOTE: Adaptec CI/O is not supported on NetWare 6.0. When upgrading a system from NetWare 5.x to NetWare 6.0, uninstall Adaptec CI/O before upgrading the operating system. ====================================================================== ISSUES FOR DIAGNOSTICS ====================================================================== The following subsections list the currently known issues regarding implementation and operation of the diagnostic tests offered in Server Administrator. ---------------------------------------------------------------------- ISSUES FOR ALL OPERATING SYSTEMS ---------------------------------------------------------------------- * After running a diagnostic test that requires a high percentage of CPU resources, you may experience slow browser response time in Netscape 6.2.x. To restore normal browser response time, you must end the Netscape 6.2.x browser process. In Red Hat Linux, locate the process ID for "netstat" and type the kill -9 command followed by the process ID number. In a Microsoft Windows operating system, open the "Task Manager" by typing , select "Netscp6.exe", and click "End Task". * When multiple diagnostics have been selected to run, the Diagnostic Progress page may show 0 percent completed for all tests. However, the diagnostics are running. Wait for the diagnostics to complete to view the results. * If the diagnostic progress shows 100 percent complete, but the "View Results" button does not appear in the upper left corner and the diagnostics will not ABORT, try relogging into Server Administrator. You might see the following message after you relogin: "Another user is currently running the diagnostics denoted by the progress bar below. When these diagnostics complete, click "Run Tests" to begin the tests you have selected". If you see this message, you must restart the Server Administrator service and run the diagnostics again. * The Diagnostic Service tab is not available in Server Administrator and is not listed on the Health page at startup. Diagnostic Service waits for the Server Administrator service to start completely before enumerating (scanning system for available devices to diagnose). If you press to refresh, the tab will show up when enumeration is complete. This process can take several minutes on some systems. * When you select "Run Tests" in the Diagnostic Selection page or in the Diagnostic Scheduler page, only some of the selected tests display in the progress page. This event occurs if you click the button very fast. This event occurs because the data you selected is being posted to the server at the same time that the page is being updated; sometimes, the interface does not update all the user changes. * If the managed system has a statically assigned IP address and no gateway is assigned to the system, then the network diagnostic communications test returns the message, "The network device must be configured to use TCP/IP and the IP for the network card must be obtained via a DHCP server. Configure the network card accordingly and run this test again". This configuration is currently not supported. * If you log in and select tests to run while tests are already running, you receive the message, "Another user is currently running the diagnostics denoted by the progress bar below. When these diagnostics complete, click "Run Tests" to begin the tests you have selected". If you leave the "Progress" page and return to the page by clicking the "View" tab, the Diagnostic Service continues to run the newly selected tests automatically after the scheduled tests have completed, even though you have not yet selected "Run Tests". The only way to stop the newly selected tests is to select "Abort" after they have begun to run. * Tests scheduled through the test scheduler or executed through the command line interface (CLI) cannot be aborted. * Aborting a diagnostic while it is being conducted might not succeed immediately if the diagnostic is performing critical operations. Abort takes place only after critical operations complete. Diagnostics that may exhibit this behavior are the AMI RAID Controller or Memory diagnostic. * If you see a warning icon notifying you that diagnostics are currently being run by another user, or if you see a blank page when attempting to access the "View" tab, you may have performed these steps: 1. Logged in to Server Administrator. 2. Navigated to the "Diagnostics"--> "View" tab. 3. Opened a command window. 4. Run diagnostics through the CLI. 5. Navigated to the "Properties"--> "Health" tab. These actions launch the Diagnostic Service page with the warning icon notifying you that diagnostics are currently being run by another user. If you then click the "Diagnostic Service" link or navigate to the "Diagnostics"--> "View" tab (while the diagnostics are still running), in Netscape 6.1, you see a blank page. In Internet Explorer, if you click the "Diagnostic Service" link, you do not receive the warning that states the diagnostics are currently being run by another user. * If your hard drive becomes full (no space left) and you run diagnostics on the managed system, Server Administrator does not show progress for the running diagnostics; instead, you see a blank page. You cannot abort the tests because the blank page is shown instead of the Progress page with the "Abort" button. To correct this issue, you must free up some space on your hard drive, reboot the server, and run the diagnostics again. * If you open multiple command prompt windows and run "omdiag" CLI commands from several different windows, the "omdiag" command might complete yet display no output or results in the command prompt window. If this event occurs, log into the Web-based GUI, select the "Diagnostics" tab, and then select the "View Progress" link on the "Diagnostic Warning" screen. * If you run the video memory diagnostic and you press a key while the diagnostic is running, the test aborts, but the result displayed is "Failed". Clicking the "Details" button on the Server Administrator home page displays a blank message box. * The memory diagnostic allocates only 2 GB of memory to perform the Memory Pattern Test on Windows NT, Windows 2000, and Red Hat Linux. ECC events are still detected on all DIMMs. * On the scheduler page, entering a date in the calendar greater than the year 2039 is not supported. * If the Network Communication Test returns the following failed result: "Invalid Device Reference: Cannot determine the configuration of the device. Verify the device driver is installed properly and the device is enabled in the Operating System. The test cannot continue", check the "Network Properties" tab. If the "IP address" value is "Not obtained" and an IP address is assigned when running ifconfig in Windows or ifconfig in Linux, then the driver is not the latest available. You must install the latest driver for the NIC from the "Server Assistant" CD before you can run diagnostics on this NIC. * The Diagnostics Service does not support Network Teaming. If you set up your NICs as a team, the Communications test will fail when run on these NICs. The message states "The network device is set up as part of a network team. You cannot run the test on this network adapter." Under certain combinations of conditions, you must restart the omawsd service by typing "omawsd restart" at the command prompt. The combination of conditions is described by a) through c) below: a) After running the diagnostic test the result that displays is a red X, indicating failure for that test result. b) When you click the details link the "Test Result:" field is blank c) The "Result Details:" field says "The test operation completed successfully." When you see the combination of conditions described in a) through c), restart the omawsd service. ---------------------------------------------------------------------- ISSUES FOR WINDOWS OPERATING SYSTEMS ---------------------------------------------------------------------- *You may see one or more of the following messages when scheduled diagnostic tests are running: "Java.exe - Unable to Locate DLL". "MemoryDiag.exe - Unable to Locate DLL". "NetDiag.exe - Unable to Locate DLL". "ScsiCtrDiag.exe - Unable to Locate DLL". The diagnostic tests that issue these messages complete as expected. However, the results of the memory, network, and SCSI controller tests are invalid. To obtain valid diagnostic test results, you must verify that Server Administrator is running in the Services Tree of your Microsoft Windows Control Panel. If Server Administrator is not running, you must restart it. To restart the service, go to the Control Panel, select the Services application, locate the Server Administrator service, right-click your mouse, and select Restart. Run the diagnostic tests again to obtain valid results. * The serial port diagnostics will fail or will not be available if you set the serial port to COM3 or COM4 in the system BIOS or by using the "BIOS Setup" tab in Server Administrator. * The parallel port diagnostics will fail or will not be available if you set the parallel port to LPT2 or LPT3 in the system BIOS or by using the "BIOS Setup" tab in Server Administrator. * For the IDE hard drive diagnostics, when clicking the "i" icon in the user interface, the location of the hard drive is listed as Port x : Master Device or Port x: Slave Device The port number for an IDE drive generally corresponds to the IDE channel in the following mapping: Port 0 = Primary Channel Port 1 = Secondary Channel Port 2 = Tertiary Channel Port 3 = Quadernary Channel However, in cases where more than two channels are present, the above relationship does not always hold. In these cases, the port number cannot be reliably correlated to a particular channel. Most often you can identify the drive that is failing by its size and/or brand. If you have trouble identifying which physical drive is failing, remove the drives from the third channel and run the tests again. If the drives pass, then one of the removed drives is the one that is failing. * The hard drive diagnostics are designed to run as a low-priority background task to limit the impact of its activity on other disk I/O activity. Because of this, it is recommended that minimal drive activity be under way while running the hard-drive diagnostics. The diagnostics may run slowly or even halt while other drive I/O is running simultaneously. The diagnostics will resume when the drive I/O has decreased. The hard drive may also fail with the following message: "Device failed a send diagnostic Command:Check SCSI cable connections and termination and run this test again. If this error still occurs, it may indicate a hardware problem". To avoid this message, run the diagnostic again when the system is not running heavy disk I/O. * During lengthy diagnostics tests when logged in to the Server Administrator home page using Microsoft Internet Explorer on a system running a supported Windows operating system, the browser might shut down unexpectedly. If this happens, press , click "Task Manager", click the browser item, and click "End Task". Log back into Server Administrator to continue diagnostics tests. * Array Manager is a requirement to diagnose RAID attached hard drives with Diagnostic Service. If Array Manager is not installed, the Elite RAID attached hard drive tests are not visible in the Diagnostic Service. * For Diagnostic Service 2.2.0, use Array Manager 3.2. If you need to upgrade to a newer version of Array Manager, then you must upgrade the Diagnostic Service to 2.2.0.A or later. The diagnostic service 2.2.0.A and 2.2.1 require Array Manager 3.3 or later. * It is recommended that no other drive activity be under way while youare running the hard drive diagnostics. The diagnostics may run slowly if other drive activity is running simultaneously. * In some configurations IDE hard drives may not be able to be tested. If the IDE hard drive is connected as a slave on the IDE bus and the master is a non-Self Monitoring and Reporting Technology (SMART) device (such as a CD drive), both the standard and extended tests fail with the following message: "The test could not be executed because the master device on the same channel as this device is not SMART-enabled. See the readme for instructions on applying fix for this issue." This problem is eliminated with Service Pack 3 for Windows 2000. See the Microsoft Knowledge Base Article ID: Q304572 at "support.microsoft.com/support/kb/articles/q304/5/72.asp." to get a hotfix for this issue. The hotfix is also located on the "Systems Management" CD in the "support\Windows\Hotfix" directory. The hotfix is in two zip files, one for Windows NT and one for Windows 2000. The files are named "MSFT_q304572_NT_SP7.zip" and "MSFT_q304572_win2k.zip". * The operating system's diskette format application and the Diagnostic Service Floppy Drive test are permitted to run simultaneously under Windows 2000. However, if you run this scenario, false diagnostic errors might occur. * If any hard drive is failing on the SCSI_INQUIRY test, then it is likely that the hard drive is not connected properly. Check the cables to ensure that the hard drive is well connected to the controller. * If you run the Adaptec RAID controller diagnostic or the Hard Disk diagnostic on hard drives that are connected to an Adaptec RAID controller with an application that reserves the controller, the diagnostic reports failure with the following message: "An application or service that manipulates RAID such as Array Manager, FAST, or PERC Console has the controller reserved and must be stopped in order to run this test. See the readme for further information". To stop the Array Manager service, go to the control panel, select the Services application, locate the Disk Management Service, right-click your mouse, and select stop. To stop the FAST or PERC Console, exit the application. * For managed systems with SCSI hard drives attached to a Dell- supported RAID controller: If the diagnostics CLI does not list the RAID hard-drive diagnostics ("raidhd" or "ideraidhd") when the Server Administrator Web server is stopped, start the Web server using the "net start Server Administrator" command; this action enables the "raidhd" or "ideraidhd" diagnostic. * When upgrading from Windows NT 4.0 to Windows 2000, you receive the following message when you install the Diagnostic Service: "Report System Compatibility: Setup checks your computer for compatibility with Windows 2000. The following items are not compatible with Windows 2000. If you continue, you may not be able to use these items, even after Setup is complete". Contrary to this message, if you continue with the upgrade, the Diagnostic Service will operate properly when the upgrade is finished. * Hard drives are sometimes incorrectly listed on the "Hardware Configuration Changed" window as being removed and then added. If the time stamp entry for both the removal and addition are identical, and you have recently upgraded the Diagnostic Service to version 2.2 or later, then the Diagnostic Service is simply detecting that information about the hard-drive controller has been added to the hardware/software description. ---------------------------------------------------------------------- ISSUES FOR RED HAT LINUX OPERATING SYSTEMS ---------------------------------------------------------------------- * The Red Hat Linux operating system may lock up when running the CD/DVD diagnostics. See the documentation about this issue at "www.exocore.com/technologies/linux/rhl71dma/". * If you have a PowerEdge 2450 system with an embedded PERC 3/Si that is set to SCSI mode in the BIOS, the SCSI controller diagnostic is not available in Online Diagnostics. * Periodically, running the command "omdiag [system | chassis] x" (where x is a diagnostic) while the Server Administrator Web server is running can cause the CLI JVM to stop and return a log file after successful completion of the diagnostic. This issue is under investigation and will be fixed in a future release. The diagnostic test completes without problems before the JVM stops; when the JVM does stop, it produces a log file. The Web server JVM is unaffected by such an event, so you can subsequently run the "omdiag" CLI command without encountering problems. * The "omreport chassis" CLI command does not list the health status of the Diagnostic Service. Diagnostic Service health can be viewed on the "Properties" page for the "Main System Chassis" object on the Server Administrator home page. * The Diagnostic Service uses the operating system to determine the number of CPUs in the system. Only CPUs that are visible to the operating system can be diagnosed. The number of CPUs listed by the Diagnostic Service equals the number of CPUs listed in the "/proc/cpuinfo" file. * The Diagnostic Service lists twice the number of physical microprocessors installed in a system if that system has Hyper-Threading enabled. The Diagnostic Service utilizes the operating system to diagnose the microprocessor. If Hyper-Threading Technology is enabled, the operating system will report twice the number of physical microprocessors installed. The "Processors Information Page" on the Server Administrator home page will list the number of physical microprocessors present on the system. If your system has Hyper-Threading Technology enabled when you test all microprocessors, the Diagnostic Service runs diagnostics on each microprocessor twice. ====================================================================== ISSUES FOR REMOTE ACCESS ====================================================================== NOTE: The Remote Access Service is supported on the PowerEdge 1600SC, 1650, 2600, 2650, 4600, 6600, and 6650 systems only. The following subsections list the currently known issues regarding implementation and operation of your RAC and the Remote Access Service in Server Administrator. ---------------------------------------------------------------------- Remote Access Dial-Up Settings ---------------------------------------------------------------------- If you change a DRAC III modem "Dial Mode" setting to "Tone" or "Pulse" using Server Administrator, IT Assistant displays the opposite setting. ---------------------------------------------------------------------- ISSUES FOR ALL OPERATING SYSTEMS ---------------------------------------------------------------------- * For the RAC to capture the last crash screen, you must enable the watchdog timer on the RAC managed system and set the watchdog recovery action to "Reset," "Power Off," or "Power Cycle." Configure the watchdog timer using Server Administrator or IT Assistant. For more information, see the "Dell OpenManage Server Administrator User's Guide," or the "Dell OpenManage IT Assistant User's Guide." * The mouse moves erratically on the screen during console redirection. The mouse movement on a local system may appear jerky and erratic during console redirection. This behavior is evident especially when using menus or opening windows. * For DRAC IIIs only, during normal operation of the DRAC III managed system under NetWare 5.1 and NetWare 6.0, the operating system periodically displays the following messages: "LCP is down: **PPP did not receive a response from peer for several Echo-Requests sent**" "Call disconnection for protocol IP to destination RAC3. The event occurred due to a link failure" These messages do not indicate errors. They are generated when the DRAC III PPP connection is temporarily unavailable (this may happen when the card is reconfigured or new users are added). The PPP connection is automatically reestablished, and no user action is required. Therefore, the user should not attempt to manually reestablish the connection. ---------------------------------------------------------------------- ISSUES FOR WINDOWS OPERATING SYSTEMS ---------------------------------------------------------------------- * In order to launch the RAC Web console from IT Assistant, you must run one of the patches located on the "Dell OpenManage Systems Management" CD in the "\support\Windows\RAC" directory. If you are running IT Assistant on any Windows operating system except Windows NT, use the "Win2KPatch.reg" file. For Windows NT, use the "WinNTPatch.reg" file. To apply the patch, browse the CD to the directory mentioned above and double-click the file's name. Click "YES" and "OK" on the next two message boxes. The patch will enable the Java Console in Internet Explorer. * On systems running Windows, you must remove any previous installation of the RAC management station software before you can install a newer version. To remove the Windows RAC management station software, perform the following steps: 1. Insert the "Systems Management CD" into the CD drive. 2. When the installation application starts, click "Exit". 3. Open a command shell window: a. Click the "Start" button, and then click "Run". b. In the "Run" dialog box, type the following line and then press : cmd 4. Start the Install Shield uninstall program to remove software: a. In the command shell window, type the following line (replace x: with the actual drive letter of your CD drive, such as d:): x:\rac13\mtpkg\setup.exe b. Select "Remove", and then click "Next". c. Click "OK". d. Click "Finish". 5. Close the command shell window. NOTE: You do not need to reboot your system. * Resizing the Windows desktop to 640 x 480 resolution causes some loss of displayed information in the browser when viewing the Server Administrator RAC home page. The behavior is the same in both Netscape Navigator and Internet Explorer. Enlarge the browser window to view all the displayed information. * The Internet Explorer application embedded in the Windows Advanced Server Limited Edition operating system does not support Java applets; therefore, it is not possible to use the Advanced Server Limited Edition's Internet Explorer browser to connect to a RAC over a local area network (LAN) or the Internet. * If you uninstall the RAC managed node software from a system running the Microsoft Windows NT operating system, remove the "RAC PPP Connection Using RACPORT" modem connection that was added during setup. Otherwise, the system displays an error message each time it reboots. * CONNECTING TO THE INTERNET IF THE RAC PPP SERVICE IS RUNNING After installing Internet Explorer 5.0 or later, you may notice that the automatic configuration and proxy setting information for each connection is separate from the same information used for the LAN connection. If you have both a dial-up and a LAN connection at the same time, Internet Explorer may block your access to the Internet. This blocking occurs because each connection in Internet Explorer 5.0 uses its own automatic configuration and proxy settings. When this has occurred and a user tries to establish any kind of PPP connection, Internet Explorer attempts to use the dial-up connection settings to access the Internet. You can find additional information about this issue in the Microsoft Knowledge Base located at "support.microsoft.com", article Q236920 entitled, "Connections No Longer Use LAN Automatic Configuration and Proxy (Q236920)." To correct this problem, perform one of the following steps: * If you are using Internet Explorer 5.0, 5.01, or 5.5 (SP1), you must update your browser to Internet Explorer 5.5 (SP2). To run the update, you need to stop the RAC service. The update is available at "windowsupdate.com". * If you are using Internet Explorer version 5.5 (SP2) or later, apply the "lanpatch.reg" registry patch by double-clicking the "lanpatch.reg" file. The patch is copied to the managed system during the installation of the DRAC III managed system software in the default director \RAC\MN. Apply the patch for each user account on the system running Internet Explorer. ###################################################################### Information in this document is subject to change without notice. (C) 2002 Dell Computer Corporation. All rights reserved. Reproduction in any manner whatsoever without the written permission of Dell Computer Corporation is strictly forbidden. Trademarks used in this text: "Dell", "PowerEdge", "PowerVault", and "Dell OpenManage" are trademarks of Dell Computer Corporation; "Microsoft", "Windows", and "Windows NT" are registered trademarks of Microsoft Corporation; "Novell" and "NetWare" are registered trademarks of Novell, Inc.; "Intel" is a registered trademark of Intel Corporation. Other trademarks and trade names may be used in this document to refer to either the entities claiming the marks and names or their products. Dell Computer Corporation disclaims any proprietary interest in trademarks and trade names other than its own. August 2002