This database stores a list of devices discovered the last time the server was operational. The device database is used to detect changes to the server. The alarm occurs when the list of current devices does not match the list stored in the device database. For example, it may occur if a hard drive is no longer operational after a power outage.
To correct this problem, delete the device database, then restart the server.
On servers running Novell NetWare, you can delete the device database in one of two ways. At the server console, either type:
del scsiflt.dev
to delete the file from the SYS:\HPSERVER directory.
Or use "c" parameter to clear the device database by typing:
unload scsiflt
load scsiflt c
On servers running Microsoft Windows NT, the device database is stored in \system32\config\SCSIFLT.DEV within the root directory of the Windows NT installation. The database is called SCSIFLT.DEV.
This alarm will not be repeated unless the server is rebooted. For more information, contact your dealer, or HP Customer Service Representative. See your User Guide for more details on technical service and support.
For more information on managing the SCSIFLT agent device database see:
SCSIFLT Agent Command Line Parameters
This information is specifically related to servers running Novell NetWare. When you manually unload the SCSIFLT file, Auto Alert displays this alarm. If you want, reload the missing NetWare Loadable Module (NLM) with the command, "load SCSIFLT". For more information on NetWare Loadable Modules, see your Novell Netware documentation.
This message indicates that a new host adapter has been installed in the server, or a device database was cleared. This message also appears upon the initial loading of the SCSI monitoring software. The alarm message "Host Adapter 0 Discovered" is displayed for this alarm, where 0 will be replaced by the host adapter ID.
No action is necessary. This message is informational only and does not indicate an error.
The SCSI host adapter is no longer detected. It may have failed, been physically removed, or improperly configured. It is also possible that the driver file for the host adapter was unloaded.
Two alarm messages are associated with this alarm. The alarm message "Host Adapter 0 Invalid" is displayed for this alarm when a command sent to the host adapter returns with an "Invalid Adapter" message. While this condition may be caused by failure of the host adapter, other causes include physically removing the adapter from the system, misconfiguring the host adapter or when the disk driver for the host adapter is not loaded. The alarm message "Host Adapter 0 Failed" is displayed when the host adapter returns a variety of operational errors. In some cases these operational errors are temporary so SCSIFLT does not report an error until two consecutive operational errors occur. With either message the 0 in the message is replaced by the appropriated host adapter ID.
Perform a graceful shutdown of the server. Close down the network operating system (NOS), and then reboot the server. Verify that the SCSI host adapter and the hard drive attached to the host adapter are responding properly. Make sure the configuration has not changed, and that it has not been corrupted. Check the cables to ensure they are properly connected. If the alarm continues, run HP Diagnostic Assistant from your HP NetServer's HP Utility Partition or the HP NetServer Navigator CD-ROM.
For more information, contact your dealer, or HP Customer Service Representative. See your User Guide for more details on technical service and support.
This message indicates a return to normal server operations. When a host adapter, which has failed, becomes accessible again, this alarm will be generated. This alarm may also be generated when a failed host adapter is replaced by an identical host adapter. The alarm message "Host Adapter 0 Recovered" is displayed for this alarm, where 0 will be replaced by the host adapter ID. No action is necessary.
Upon server startup, the identification information received from the SCSI host adapter is compared with the corresponding data in the device database. If the identification information in the database no longer matches the host adapter information, this alarm is generated. If an adapter is replaced with an identical adapter, no change will be detected. The alarm message "Host Adapter 0 Changed" is displayed for this alarm, where 0 will be replaced by the host adapter ID.
No action is necessary. This message is informational only and does not indicate an error.
This alarm is usually caused by the failure of the SCSI host adapter associated with the disk driver. In some cases, it may be caused by a failure of the operating system. Regardless of the cause, this alert usually indicates an unrecoverable error. The alarm message "Host Adapter 0 Timeout" is displayed for this alarm, where 0 will be replaced by the host adapter ID.
Perform a graceful shutdown of the server. Close down the network operating system (NOS), and then reboot the server. Verify that the SCSI host adapter and the hard drive attached to the host adapter are responding properly. Make sure the configuration has not changed, and that it has not been corrupted. Check the cables to ensure they are properly connected. If the alarm continues, run HP Diagnostic Assistant from your HP NetServer's HP Utility Partition or the HP NetServer Navigator CD-ROM.
For more information, contact your dealer, or HP Customer Service Representative. See your User Guide for more details on technical service and support.
This message indicates that a new SCSI device has been installed in the server, such as a CD-ROM player, hard disk drive, or tape drive is first discovered, or a device database was cleared. This message also appears upon the initial loading of the SCSI monitoring software. The alarm message "Device Hard Disk 0::1 Discovered" is displayed for this alarm. The type of device will replace "Hard Disk" where appropriate, and the 0::1 will be replaced by the host adapter ID and device ID, in that order.
No action is necessary. This message is informational only and does not indicate an error.
A SCSI device installed in the server could not be detected. The failure to detect the device may be caused by a power or cable problem, or a change in the server configuration. In the alarm message, "Device Hard Disk 0::1 Failed," the type of device that failed replaces "Hard Disk" where appropriate. The "0::1" will be replaced by the host adapter ID and device ID, in that order.
Perform a graceful shutdown of the server. Close down the network operating system (NOS), and then reboot the server. Verify that the SCSI device and the hard drive attached to the device are responding properly. Make sure the configuration has not changed, and that it has not been corrupted. Check the cables to ensure they are properly connected. If the alarm continues, run HP Diagnostic Assistant from your HP NetServer's HP Utility Partition or the HP NetServer Navigator CD-ROM.
For more information, contact your dealer, or HP Customer Service Representative. See your User Guide for more details on technical service and support.
This message indicates a return to normal server operations. It also appears when a failed SCSI device is replaced with an identical device. In the alarm message, "Device Hard Disk 0::1 Recovered," the type of device will replace "Hard Disk" where appropriate. The "0::1" will be replaced by the host adapter ID and device ID, in that order.
No action is necessary.
Upon server startup, the identification information received from the SCSI device is compared with corresponding values in the device database. If the identification information in the database no longer matches the device information, this message is generated. If a device is replaced with an identical device, no change will be detected. The alarm message "Device Hard Disk 0::1 Changed" is displayed for this alarm. The type of device will replace "Hard Disk" where appropriate, and the 0::1 will be replaced by the host adapter ID and device ID, in that order.
No action is necessary. This message is informational only and does not indicate an error.
This alarm is usually caused by the failure of the disk driver associated with the SCSI host adapter to which the device is attached. In some cases, it may be caused by a failure of the operating system. Regardless of the cause, this alarm usually indicates an unrecoverable error. The alarm message "Device Hard Disk 0::1 Timeout" is displayed for this alarm. The type of device will replace Hard Disk where appropriate, and the 0::1 will be replaced by the host adapter ID and device ID, in that order.
Perform a graceful shutdown of the server. Close down the network operating system (NOS), and then reboot the server. Verify that the SCSI device and the hard drive attached to the device are responding properly. Make sure the configuration has not changed, and that it has not been corrupted. Check the cables to ensure they are properly connected. If the alarm continues, run HP Diagnostic Assistant from your HP NetServer's HP Utility Partition or the HP NetServer Navigator CD-ROM.
For more information, contact your dealer, or HP Customer Service Representative. See your User Guide for more details on technical service and support.
The following command line parameters are supported for the SCSIFLT agent file on the server.
|
Parameter |
Description |
|
A |
This will send a trap when the NLM is unloaded or the database is corrupted. |
|
Fn |
This sets the polling frequency in n minutes (example: F60 tells the agent to poll once per hour). |
|
C |
This clears the device database. |
If you need to delete the database file manually, the database location is: SYS:\HPSERVER and is called SCSIFLT.DEV.
|
Parameter |
Description |
|
A |
This is not applicable. |
|
Fn |
This sets the polling frequency in n minutes (example: F60 tells the agent to poll once per hour). |
|
C |
This clears the device database. |
If you need to delete the database file manually, the database location is in the directory where the HP NetServer SNMP agents have been installed and is called SCSIFLT.DEV.
For Windows NT, registry entries are used in place of command line parameters. Parameters are put into the Windows NT registry under the following key:
HKEY_LOCAL_MACHINE
SOFTWARE
ADAPTEC
SNMP_SCSI_ExtensionsAgent
Command_Line_Parameters
You may add one or more of the following values:
|
Value |
Description |
|
1:REG_SZ:AdminAlert |
This enables traps (remove value to disable traps). |
|
2:REG_SZ:ClearDevFile |
This clears the device database. |
|
4:REG_SZ:Frequency n |
This sets the polling frequency to n minutes (example: 4:REG_SZ:Frequency 60 tells the agent to poll once per hour). |
If you need to delete the database manually, the database location is in the \system32\config directory under the root of your Windows NT installation and is called SCSIFLT.DEV.
|
Parameter |
Description |
|
-a |
This will send a trap when the scsiflt agent is unloaded or the database is corrupted. |
|
-f n |
This sets the polling frequency in n seconds (example: f 60 tells the agent to poll once per minute). |
|
-c |
Starts the scsiflt agent and Invokes the SCSI device database installation program (SFInstall) to allow you to add or remove devices. |
If you ever add or remove SCSI devices from your server you must manually update the SCSI device database by running the SFInstall program. It can be run interactively from the command line by giving the full path:
/usr/lib/hpnsa/scsiflt/sfinstall
or by invoking the SCSIFLT SNMP Peer daemon (/usr/lib/hpnsa/scsiflt/scsiflt) with the -c option (see above).
The SFInstall program has the following command line parameters:
|
Parameter |
Description |
|
-c |
Unconditionally rebuilds the SCSI device database and interactively allows editing of the device list without starting scsiflt. |
|
-d n |
This runs the SFInstall program with the debugging level set to n. Levels are 0 (off) to 9 (excessive). |