Basic TroubleshootingDell™ Remote Access Controller Installation and Setup Guide
Troubleshooting Network Problems
The purpose of this section is to help you diagnose and possibly solve problems that may occur with your RAC.
This section provides solutions to common problems.
Console redirection does not work.
If you do a hardware reset on the RAC without rebooting the managed system, the console redirection stops working and erases the RAC configuration space in RAC memory. A software reset does not cause this problem. To solve this problem, you must reboot the managed system.
The console redirect frame shows Please wait - initial screen loading, and seems to hang in this mode when the managed system is up, and VNC and PPP are running on the managed system.
No connection may exist between the managed system and the firmware through a PPP connection. Rebooting the managed system may correct this problem. It is not necessary to reboot the DRAC III.
Performing a hard reset of the RAC firmware without rebooting the managed system causes the loss of the connection to the console redirection service.
Performing a hard reset of the RAC firmware without rebooting the managed system causes console redirection to stop working in graphic mode.
 |
NOTE: Graphics console redirection stops only after a hard reset of the RAC firmware; this does not occur after a soft reset. Server Administrator performs a soft reset of the RAC firmware after performing configuration actions such as adding a user to the RAC. It is not necessary to reboot the managed node under these circumstances. |
You may see the following message on the bottom menu of the console redirect frame:
Please wait - initial screen loading.
You cannot correct the problem by closing and restarting the console redirect frame. You must reboot the managed system to reestablish the connection to the console redirection service to correct the problem.
The redirected console screen does not refresh.
Click Refresh in the console redirection window.
The following message is displayed: Warning: remote console is not available.
This warning indicates one of the following conditions:
After waiting a few minutes for the system to restart or the screen to change modes, try restarting the console redirect window to correct the problem.
The IPMI interfaces do not provide the correct information.
On Dell™ PowerEdge™ 1650 systems, the DRAC III is installed on a riser board. The riser board plugs into the RISER connector on the system board and is considered an extension of the system board. There are two riser board configurations for the PowerEdge 1650. The first features two 64-bit, 66-MHz expansion slots. The second features one 64-bit, 66-MHz expansion slot (PCI2) and one 32-bit, 33-MHz expansion slot (PCI1) for 5-V cards. In PowerEdge 1650 systems, if the DRAC III is installed on a riser card equipped with two 64-bit slots, the card operates, but the IPMI interface information provided is incorrect. To correct this problem, ensure that the DRAC III is installed on a 5-V riser card equipped with one 32-bit slot. For more information, see "Installing a DRAC III or ERA/O."
Text redirection is not occurring when using console redirection.
This situation could occur if the DRAC III is not installed on the primary PCI bus. The DRAC III must be installed in PCI slot number 1 to display screen text activity.
Ensure that the DRAC III is installed on the primary PCI bus. For more information, see "Installing a DRAC III."
Graphics redirection is not occurring when using console redirection.
This situation could occur if the RAC console redirection service is not installed properly or is not running.
After allowing several minutes for the graphics redirection to occur, ensure that the RAC console redirection service is running. Try stopping and then starting the service. If the problem persists, reboot the system.
The internal RAC Trace Log can be used by administrators needing to debug alerting, paging, or networking from the RAC. The Trace Log can be accessed from the RAC Web-based remote access interface by clicking the Debug tab, and then clicking Network Debug. From the Network Debug window, select Dump Trace Log, and then click Submit. The Trace Log tracks the following information:
|
NOTE: Settings for CHAT, DHCP, IP (DRAC III only), PPP (DRAC III only), and TAP (DRAC III only) can be accessed from the RAC remote access GUI by clicking the Debug tab, and then clicking Trace Level. |
|
NOTE: In the RAC Trace Log, nonprintable ASCII characters are translated to printable ASCII characters. If the character code is less than 0x20, or between 0x7f and 0xa0 (inclusive), the value 0x40 is exclusive-or'd with the character before printing, after a "^" is added to the beginning. Thus, the ASCII carriage return character, 0xd, is printed as "^M" in the Trace Log. Nonprintable ASCII characters may occur during tracing of the CHAT and TAP protocols, and occasionally during PPP negotiations. |
Some paging services return a busy signal when a paging request is successfully accepted. This cannot be distinguished from the case where the line is busy, and the paging service never answered. Therefore, even though the chat script expects BUSY, this is indicated as a failure on the trace log.
A chat script time-out is considered a success indication for numeric paging, because no other error indications were detected. Since numeric paging services do not have a positive confirmation indication that can be detected by the modem, numeric paging is inherently unreliable. For this reason, up to three numeric paging attempts are made, and duplicate numeric pages may be received.
The RAC provides a standard set of network diagnostic tools, similar to those found on Windows® or Red Hat Linux-based systems. Using the RAC Web-based remote access interface, you can access the following network debugging tools by clicking the Debug tab and then clicking Network Debug. For more information about the Network Debug feature, see the remote access interface help.
The trace log may also contain RAC operating-system specific error codes (relating to the internal RAC operating system, not the managed system's operating system). Table D-1 can help you diagnose network problems reported by the internal RAC operating system:
|
Error Code |
Description |
|---|---|
0x5006 | ENXIO: No such address. |
0x5009 | EBADS: The socket descriptor is invalid. |
0x500D | EACCESS: Permission denied. |
0x5011 | EEXIST: Duplicate entry exists. |
0x5016 | EINVALID: An argument is invalid. |
0x5017 | ENFILE: An internal table has run out of space. |
0x5020 | EPIPE: The connection is broken. |
0x5023 | EWOULDBLOCK: The operation would block; socket is nonblocking. |
0x5024 | EINPROGRESS: Socket is nonblocking; connection not completed immediately. |
0x5025 | EALREADY: Socket is nonblocking; previous connection attempt not complete. |
0x5027 | EDESTADDRREQ: The destination address is invalid. |
0x5028 | EMSGSIZE: Message too long. |
0x5029 | EPROTOTYPE: Wrong protocol type for socket. |
0x502A | ENOPROTOOPT: Protocol not available. |
0x502B | EPROTONO SUPPORT: Protocol not supported. |
0x502D | EOPNOTSUPP: Requested operation not valid for this type of socket. |
0x502F | EAFNOSUPPORT: Address family not support. |
0x5030 | EADDRINUSE: Address is already in use. |
0x5031 | EADDRNOTAVAIL: Address not available. |
0x5033 | ENETUNREACH: Network is unreachable. |
0x5035 | ECONNABORTED: The connection has been aborted by the peer. |
0x5036 | ECONNRESET: The connection has been reset by the peer. |
0x5037 | ENOBUFS: An internal buffer is required but cannot be allocated. |
0x5038 | EISCONN: The socket is already connected. |
0x5039 | ENOTCONN: The socket is not connected. |
0x503B | ETOOMANYREFS: Too many references, cannot splice. |
0x503C | ETIMEDOUT: Connection timed out. |
0x503D | ECONNREFUSED: The connection attempt was refused. |
0x5041 | EHOSTUNREACH: The destination host could not be reached. |
0x5046 | ENIDOWN: NI_INIT returned -1. |
0x5047 | ENMTU: The MTU is invalid. |
0x5048 | ENHWL: The hardware length is invalid. |
0x5049 | ENNOFIND: The route specified cannot be found. |
0x504A | ECOLL: Collision in select call; these conditions already selected by another task. |
0x504B | ETID: The task ID is invalid. |
Use the following information to troubleshoot a particular type of RAC alert:
RAC Log messages can be used by administrators to debug alerting from the RAC. Table D-2 provides a list of RAC log message IDs, message and description, as well as corrective actions to take for a particular message.
|
NOTE: In Table D-2, the character "L" is sometimes used in the Message ID column. "L" represents the severity level or type of the message, which can be one of the following: W (warning), E (error), S (severe), F (fatal), or A (always). |
The DRAC III has two LEDs located on the back of the card connector. The top LED is green, and is called the heartbeat LED. The amber LED is below the green, and is called the error LED.
The following are conditions indicated by the DRAC III LEDs:
If the amber LED is solid, it indicates a nonrecoverable error. A nonrecoverable error occurs when a POST memory test or core operation has failed, and the DRAC III cannot proceed with a boot process. The DRAC III must be replaced.
Summary for this condition:
If the amber LED is flashing at .5-second intervals, it indicates that the core, firmware, database, or production sector in the DRAC III flash is corrupted. A field technician must replace the DRAC III.
Summary for this condition:
The following sections define the blink codes that are produced by the amber error LED if an error is detected by any of the self-tests or extended self-tests.
The blink code repeats about every 10 seconds. For example, a code of 3114 (a problem in the uart loopback test) causes the amber LED to flash three times, pause, flash one time, pause, flash one time, pause, flash four times. The sequence then repeats after 10 seconds.
1111 = Unable to create a self-test task.
1112 = A self-test task is currently running. (Multiple self-tests cannot be started.)
1113 = Failure creating a self-test visual signal.
1114 = Failure to allocate required DRAC III system memory.
1115 = Failure writing the D_selftest_BDSTATUS.
1116 = Failure attempting to send a debug message.
1117 = Error when accessing the DRAC III database.
2111 = Failure in extended memory testing — Read verify, write.
2112 = Failure in extended memory testing — Read verify write high memory to low.
2113 = Failure in extended memory testing — Read verify, write, write, low-to-high.
2114 = Failure in extended memory testing — Read verify, write, write, high-to-low.
2115 = Failure in extended memory testing — Read verify, high-to-low.
2116 = Failure in extended memory testing — Read verify, low-to-high.
2117 = Failure in marching memory test — Read verify, write.
2118 = Failure in marching memory test — Read verify in low-to-high memory.
3111 = Failure opening uart for external loopback.
3112 = Failure in I/O control to uart driver.
3113 = Failure writing data to the uart.
3114 = Failure reading data from the uart.
3115 = Transmit/receive data miscompare.
3116 = Failure trying to suspend VT-100 task.
4111 = Failure in the GPIO green LED test.
4112 = Failure in the GPIO LED test.
4113 = SMI connector GPIOs not reading inactive values.
5111 = More than one power source is selected (internal DRAC III problem).
5112 = No power source is shown to be driving (internal DRAC III problem).
5113 = Failure in the onboard hardware monitor sensors/logic. (The managed system must be powered up or the PCI voltage tests fail.)
5114 = Failure accessing data base for hardware monitor parameters.
5115 = Failure in accessing the onboard hardware monitor.
5121 = DRAC III battery voltage is out of range.
5122 = DRAC III external power adaptor voltage is out of range.
5123 = PCI AUX 3.3 voltage is out of range.
5124 = PCI +5 voltage is out of range.
5125 = PCI -12 voltage is out of range.
5126 = PCI +12 voltage is out of range.
5127 = DRAC III temperature monitor is out of range.
5128 = DRAC III battery presence is not detected.
5129 = DRAC III external power adaptor presence is not detected.
6111 = No IPMI connector is detected.
6112 = IPMI Get Chassis Status command to the BMC failed.
7111 = Failure when loading the EXPROM image from the database into shared memory.
7112 = Failure when loading the EXPROM header from the database into shared memory.
7113 = Invalid EXPROM header signature.
7114 = Invalid EXPROM vendor or device ID.
8111 = Failure erasing U16 (Firmware) diagnostic sector.
8112 = Failure writing U16 (Firmware) diagnostic sector.
8113 = Failure read/verify U16 (Firmware) diagnostic sector.
8114 = Failure erasing U17 (DataBase) diagnostic sector.
8115 = Failure writing U17 (DataBase) diagnostic sector.
8116 = Failure writing U17 (DataBase) diagnostic sector.
9111 = Failure in PCMCIA to DRAC III interface.