Monitoring and Optimizing the Server

This section describes how to prevent server problems that involve memory management, CPU utilization, process scheduling, and LAN traffic. There are also procedures to help improve server performance.


Assessing Server RAM

One of the most common causes of a slow network is insufficient Random Access Memory (RAM) in the NetWare server. This procedure helps you determine if your server has enough RAM.


Procedure

  1. At the server console prompt, type

    LOAD MONITOR <Enter>

  2. Choose Cache Utilization from the Available Options menu.

  3. Write down the percentage of Long Term Cache Hits shown in the lower window of the screen.

    Tracking cache hits helps you predict when you need more RAM before you experience serious performance loss.

    If the percentage of Long Term Cache Hits falls below 90 percent, add more RAM to the server.

  4. (Optional) To temporarily free up RAM, unload any NLM programs that are not critical.

  5. To exit MONITOR, press <Esc>, or to return to the console prompt without exiting MONITOR, press <Alt>+<Esc>.


Additional Information

For more information about See

Cache buffers

Cache buffer in Concepts

MONITOR utility

MONITOR in Utilities Reference


Controlling Memory Allocation

This procedure explains how to limit the amount of memory available for file and directory caching, thus making more memory available for NetWare Loadable ModuleTM (NLM) programs and server processes. Use this procedure if you receive insufficient memory alerts when trying to load NLMTM programs.

To solve memory problems without limiting memory for file and directory caching, add more memory.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose File Caching from the Categories menu.

  4. Choose Minimum File Cache Buffers.

    This parameter specifies the minimum amount of memory the system will keep for file caching. Memory above this minimum amount is available for other processes.

  5. Reduce the value of this parameter.

  6. Press <Esc> twice to reach the Categories menu.

  7. Choose Directory Caching from the Categories menu.

  8. Choose Maximum Directory Cache Buffers.

    This parameter specifies the maximum number of buffers that can be allocated for directory caching. Once allocated, the buffers remain allocated, even if they are not being used. Thus, they are not available for other processes.

    By reducing this parameter, you prevent the system from allocating too many directory cache buffers and you ensure that more memory is available for other server processes.

    NOTE:  Reducing the maximum number of directory cache buffers may impede system performance.

  9. Reduce the value of the Maximum Directory Cache Buffers parameter.

  10. Press <Esc> twice to reach the Update Options menu.

  11. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  12. Press <Enter> to update the file.

    When you press <Enter>, the system writes the parameters to the AUTOEXEC.NCF file or updates the parameters if they are already in the file.

  13. Reboot the server to free the memory.


Speeding Directory Searches

Use this procedure if the server responds slowly to directory searches.

NOTE:  This procedure requires you to increase the number of directory cache buffers that can be allocated by the system. Memory allocated for directory cache buffers is no longer available for use by other processes.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose Directory Caching from the Categories menu.

  4. Select Directory Cache Allocation Wait Time.

    This parameter specifies how long the system must wait after allocating one directory cache buffer before it can allocate another buffer. If the value is too high, directory searches are slow.

  5. Decrease the value of this parameter.

  6. Select Maximum Directory Cache Buffers.

    This parameter specifies the maximum number of cache buffers the system can allocate for directory caching. If the number is too low, directory searches are slow.

  7. Increase the value of this parameter.

  8. Select Minimum Directory Cache Buffers.

    This parameter specifies the minimum number of directory cache buffers the system allocates. If the number is too low, directory searches are slow.

  9. Increase the value of this parameter.

  10. Press <Esc> twice to reach the Update Options menu.

  11. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  12. If desired, press <Enter> to update the file.

    If you press <Enter>, the system writes the parameters to the AUTOEXEC.NCF file or updates the parameters if they are already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter changes last only until you reboot the server.


Prioritizing Server Processes

You can use SCHDELAY to prioritize server processes. SCHDELAY allows you to delay the execution of specific processes, thereby giving more CPU time---and higher priority---to other processes.

The command syntax is

LOAD SCHDELAY process_name = number [process_name = number] <Enter>

The process name value is the name of a process or thread that accesses the server CPU. You can specify the process value for multiple processes at one time.

To view the list of processes that can be prioritized, along with their current SCHDELAY values, type

LOAD SCHDELAY <Enter>

The number value determines how many times the process delays execution. A value of 0 causes no delay, while a value of 2 skips every other execution of the process.

To delay a process, you can enter a value from 2 to 10000---the higher the value, the lower the priority of the process.

NOTE:  The SCHDELAY values only have an effect if the system is busy and two or more processes are vying for the CPU's time.

For example, if you want to assign a database NLM called DBNLM a lower priority than other processes so it won't monopolize the NetWare server's CPU, put a command similar to the following in the AUTOEXEC.NCF file:

LOAD SCHDELAY DBNLM = 2

To view the list of processes and determine which processes are hoarding CPU time, follow this procedure.


Procedure

  1. At the server console prompt, type

    LOAD MONITOR <Enter>

  2. Select Scheduling Information from the Available Options menu.

    A list of server processes appears, with four columns of information about each process.

    • The Sch Delay column shows the current SCHDELAY value assigned to the process. The default is 0.
    • The Time column shows the CPU time spent executing the process code.
    • The Count column shows how many times the process ran during the sample period. (The Total Sample Time is displayed at the bottom of the screen.)
    • The Load column shows the percent of total CPU time spent on the process.

  3. Select a process that has a high number in the Load column.

    Look for processes that have consistently high values, rather than high peak values.

  4. Use the plus (+) or minus (-) key to increase or decrease the SCHDELAY value for the selected process.

    The value in the Load column changes accordingly.

  5. Experiment with higher or lower SCHDELAY values until the CPU load, as displayed in the Load column, is within acceptable limits.

    What is acceptable depends upon your system and your needs. The goal is to balance the load so that processes run to your satisfaction.

    NOTE:  To make the SCHDELAY value effective each time the server boots, put the SCHDELAY command in the AUTOEXEC.NCF file.

  6. To exit the scheduling information window, press <Esc>.

    You do not need to reboot the server.


Checking a Server's Processor Utilization

One possible cause of network problems is the monopolizing of the CPU by a single process. This procedure helps you determine which processes are using the CPU and how much load they are putting on the CPU's time.

For information about processor utilization in a multiprocessing server, see Maintaining NetWare SMP


Procedure

  1. At the server console prompt, type

    LOAD MONITOR <Enter>

  2. Choose Processor Utilization from the Available Options menu.

  3. Press <F3> to view a list of the currently loaded processes and interrupts.

    For each item on the list, notice the percentage given in the Load column. This is the percentage of time the CPU spent servicing this process during the one-second sample period.

    Compare processes to determine which processes are taking the most CPU time.

  4. (Optional) To redistribute processor utilization, you can do either or both of the following:


Viewing a Server's Packet Traffic

You can view the packet traffic coming in and going out of the server by using TRACK ON at the server console prompt. Use TRACK OFF to stop viewing the packet traffic.

See TRACK ON and TRACK OFF in Utilities Reference for more information about these commands.


Viewing LAN Driver Statistics

Reviewing the LAN driver statistics on your server can help you regulate network traffic and troubleshoot system problems.

For example, you can use the LAN driver statistics to determine

The number and kinds of statistics available depend on the LAN drivers installed on your system. Follow this procedure to view the statistics.


Procedure

  1. At the server console prompt, type

    LOAD MONITOR <Enter>

  2. Select LAN/WAN Information from the Available Options menu.

  3. Select a LAN driver from the LAN Driver Information menu.

    The statistics for the selected LAN driver are displayed. Press <PageUp> and <PageDown> to scroll through the information in the window. For an explanation of the statistics, see Appendix A, LAN Driver Statistics, of Utilities Reference.


Improving Server Performance

You can increase server performance two ways:

NOTE:  Use this procedure only if the network boards in your server can transmit more than 512 bytes of data per packet. Refer to the documentation that came with the network boards to determine the packet size.


Prerequisites


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose Communications from the Categories menu.

  4. Choose Maximum Physical Receive Packet Size from the Communications menu.

  5. Increase the value and press <Enter>.

    Refer to the documentation that comes with the boards to determine the maximum value.

  6. Press <Esc> twice to reach the Update Options menu.

  7. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the STARTUP.NCF file.

  8. Press <Enter> to update the file.

  9. When you want the changes to take effect, reboot the server.


Improving Disk Writes

The following sections describe methods for improving the speed of disk writes.


Increasing the Number of Concurrent Writes

You can increase the speed and efficiency of disk cache writes by increasing the number of write requests that can be executed at one time.

To determine if you need to increase the number of concurrent writes, first compare the number of dirty cache buffers to the total number of cache buffers. Dirty cache buffers contain data that has not yet been written to disk. These statistics are found on the General Information screen in MONITOR.

If the number of dirty cache buffers is greater than 70 percent of total cache buffers, use the following procedure to increase the number of concurrent write requests.

For an explanation of file caching and dirty cache buffers, refer to Cache memory in Concepts.

NOTE:  Increasing the number of concurrent disk cache writes slows disk cache reads.You will want to balance the speed of disk writes and reads to meet your own needs.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose File Caching from the Categories menu.

  4. Choose Maximum Concurrent Disk Cache Writes from the File Caching menu.

  5. Increase the parameter value and press <Enter>.

    If the parameter is currently at the default value of 50, try increasing it to 100.

  6. Press <Esc> twice to reach the Update Options menu.

  7. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  8. If desired, press <Enter> to update the file.

    If you press <Enter>, the system writes the parameter to the AUTOEXEC.NCF file or updates the parameter if it is already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter change lasts only until you reboot the server.


Changing Disk and Directory Caching for Faster Writes

Use this procedure if network users frequently make many small write requests and the server is slow to respond to the requests.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose File Caching from the Categories menu.

  4. Choose Dirty Disk Cache Delay Time from the File Caching menu.

  5. Increase the value and press <Enter>.

    This parameter specifies how long the system waits before writing a not-completely-dirty cache buffer to disk.

    If the value is low, the system writes to disk more frequently, but writes fewer requests each time. If the value is high, the system waits longer before writing to disk, but executes more write requests with each operation. A higher value provides greater efficiency in writing to disk.

    If the parameter is currently at the default value of 3.3 seconds, try increasing the value to 7 seconds.

  6. Press <Esc> to reach the Categories menu.

  7. Choose Directory Caching from the Categories menu and press <Enter>.

  8. Choose Dirty Directory Cache Delay Time.

    This parameter specifies how long the system keeps a directory table write request in memory before writing it to disk.

    Increasing the parameter provides slightly faster performance, but may increase the chance of directory tables becoming corrupted.

  9. Increase the value and press <Enter>.

    If the parameter is currently at the default value of 0.5 second, try increasing the value to 2 seconds.

  10. Choose Maximum Concurrent Directory Cache Writes.

    This parameter determines how many write requests from directory cache buffers are executed at one time. Increasing this value increases the efficiency of directory cache write requests.

    NOTE:  Increasing the number of concurrent directory cache writes decreases the speed of directory cache reads. You will want to balance the speed of writes and reads to meet your own needs.

  11. Increase the value and press <Enter>.

    If the parameter is currently at the default value of 10, try increasing the value to 25.

  12. Press <Esc> twice to reach the Update Options menu.

  13. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  14. If desired, press <Enter> to update the file.

    If you press <Enter>, the system writes the parameters to the AUTOEXEC.NCF file or updates the parameters if they are already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter changes last only until you reboot the server.


Turning Off Read-After-Write Verification

Read-after-write verification is an important means of protecting the data on your system. Normally, you should not disable it.

However, if your disks are mirrored and reliable, you may choose to disable read-after-write verification because disabling almost doubles the speed of disk writes.

WARNING:  Turning off read-after-write verification may increase the risk of data corruption on the server's hard disk. You should use the following procedure only if your disks are mirrored and reliable, and you understand the risk.


Prerequisites


Procedure

  1. At the server console prompt, type

    LOAD MONITOR <Enter>

  2. Choose Disk Information from the Available Options menu and press <Enter>.

  3. Select a disk drive from the System Disk Drives menu.

  4. Select Read After Write Verify from the Driver Status menu.

  5. Select Disable Verify from the Read After Write Status menu.

    NOTE:  You can also disable read-after-write verification by setting the Enable Disk Read After Write Verify SET parameter to OFF. However, this setting only affects those disks loaded after the parameter value is changed. It does not change the setting for currently loaded disks.


Improving Disk Reads

The following sections describe methods for improving the speed of disk reads.


Changing Concurrent Disk and Directory Writes for Faster Reads

Use this procedure if your server is slow to respond to read requests.

NOTE:  This procedure requires you to decrease the values of the Maximum Concurrent Disk Cache Writes parameter and the Maximum Concurrent Directory Cache Writes parameter. Although decreasing these values increases the speed of read requests, it may decrease the speed and efficiency of write requests.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose File Caching from the Categories menu.

  4. Choose Maximum Concurrent Disk Cache Writes.

  5. Decrease this value and press <Enter>.

    If the parameter is currently set to the default value of 50, try setting the value to 10.

  6. Press <Esc> to reach the Categories menu.

  7. Choose Directory Caching from the Categories menu.

  8. Choose Maximum Concurrent Directory Cache Writes.

  9. Decrease this value and press <Enter>.

    If the parameter is currently set to the default value of 10, try setting the value to 5.

  10. Choose Directory Cache Buffer NonReferenced Delay.

    This parameter specifies how long a directory entry must be cached before it can be overwritten. Increasing this value causes the system to allocate more directory cache buffers and thus speeds directory access.

  11. Increase this value and press <Enter>.

    If the parameter is currently set to the default value of 5.5 seconds, try setting the value to 60 seconds.

  12. Press <Esc> twice to reach the Update Options menu.

  13. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  14. If desired, press <Enter> to update the file.

    If you press <Enter>, the system writes the parameters to the AUTOEXEC.NCF file or updates the parameters if they are already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter changes last only until you reboot the server.


Changing the Turbo FAT Wait Time for Faster Reads

When a program randomly accesses a file that contains more than 64 File Allocation Table (FAT) entries, the system builds a turbo FAT index for the file so that the information in the file can be accessed quickly.

The Turbo FAT Re-Use Wait Time parameter specifies how long a turbo FAT index remains in memory after the indexed file is closed. When the turbo FAT index is in memory, files can be opened and information accessed faster.

If network users frequently access files larger than 64 blocks, use this procedure to increase the time the index is kept in memory.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose File System from the Categories menu.

  4. Select Turbo FAT Re-Use Wait Time.

  5. Increase this value and press <Enter>.

    You must specify the value in seconds.

    If the parameter is currently set to the default value of 329.5 seconds (5 minutes 29.6 seconds), try setting the value to 600 seconds (10 minutes).

  6. Press <Esc> twice to reach the Update Options menu.

  7. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  8. If desired, press <Enter> to update the file.

    If you press <Enter>, the system writes the parameters to the AUTOEXEC.NCF file or updates the parameters if they are already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter changes last only until you reboot the server.


Controlling Resource Allocation with Locks

A lock prevents a file or record from being updated by more than one user at a time. By controlling the number of file and record locks available to a workstation or a server, you control access to files and records.

You may choose to limit the number of locks to prevent overuse of file resources. Or you may choose to increase the number of locks if workstations can't open files.


Keeping Workstations from Overusing File Resources

If workstations are overusing file resources by opening and locking too many files or records at one time, use the following procedure to limit the number of locks for the server and for each workstation.

Before limiting file and record locks, make sure the workstations have an adequate number of locks to access the files and records they need. Limiting needed locks can cause applications to generate errors.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose Locks from the Categories menu.

  4. Decrease the following parameter values to limit the total number of file and record locks allowed for the server:

    Maximum Record Locks

    Maximum File Locks

  5. Decrease the following parameter values to limit the total number of file and records locks allowed for each workstation:

    Maximum Record Locks Per Connection

    Maximum File Locks Per Connection

  6. Press <Esc> twice to reach the Update Options menu.

  7. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  8. If desired, press <Enter> to update the file.

    If you press <Enter>, the system writes the parameters to the AUTOEXEC.NCF file or updates the parameters if they are already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter changes last only until you reboot the server.


Increasing File Record Locks

Use this procedure if applications fail because they cannot open enough files or because not enough record locks are available.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose Locks from the Categories menu.

  4. Increase the following parameter values to increase the number of file and record locks allowed for the server:

    Maximum Record Locks

    Maximum File Locks

  5. Increase the following parameter values to increase the number of file and record locks allowed for each workstation:

    Maximum Record Locks Per Connection

    Maximum File Locks Per Connection

  6. Press <Esc> twice to reach the Update Options menu.

  7. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  8. Press <Enter> to update the file.

    If you press <Enter>, the system writes the parameters to the AUTOEXEC.NCF file or updates the parameters if they are already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter changes last only until you reboot the server.


Displaying Alert Messages

You can use SET parameters to control the display of alert messages on the server console.


Turning on Hardware and NLM Alerts

To display alerts about hardware devices and NLM programs, add the following SET commands to the beginning of the STARTUP.NCF file:

SET Display Spurious Interrupt Alerts=ON

SET Display Lost Interrupt Alerts=ON

SET Display Relinquish Control Alerts=ON

SET Display Old API Names=ON

To temporarily turn off an alert message display (when it is filling up the console screen, for example), use the SET command to change the relevant parameter setting to OFF. When you reboot the server, the setting is changed to ON again.


Turning on Early Warning Alerts for Cache Buffer Memory Allocation

All memory not allocated for other processes is given to file caching. As memory is requested for other processes, the server gives up file cache buffers. However, there is a minimum number of file cache buffers that cannot be given up. This minimum is specified by the Minimum File Cache Buffers SET parameter.

Use the following procedure to specify when the system will warn you that the number of buffers is approaching the minimum level.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose File Caching from the Categories menu.

  4. Select Minimum File Cache Buffers.

    View this parameter to learn the minimum number of file cache buffers currently set.

  5. Choose Minimum File Cache Report Threshold from the File Caching menu.

    This parameter specifies how many unused cache buffers must be left above the minimum value for the system to warn you that cache buffers are low.

  6. Change the parameter value as needed and press <Enter>.

    For example, if the Minimum File Cache Buffers parameter is set to 20 and the Minimum File Cache Report Threshold is set to 25, you are warned when all but 45 cache buffers are allocated for other processes.

  7. Press <Esc> twice to reach the Update Options menu.

  8. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  9. If desired, press <Enter> to update the file.

    If you press <Enter>, the system writes the parameter to the AUTOEXEC.NCF file or updates the parameter if it is already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter change lasts only until you reboot the server.


Turning on Early Warning Alerts for a Full Volume

Use this procedure to enable early warning that a volume is nearly full.

If a volume is becoming full, you may need to remove unnecessary files or add a new hard disk.


Procedure

  1. At the server console prompt, load either SERVMAN or MONITOR.

    LOAD SERVMAN | MONITOR <Enter>

    NOTE:  You can also set parameters with the SET command. See SET in Utilities Reference.

  2. Choose Server Parameters from the Available Options menu.

  3. Choose File System from the Categories menu.

  4. Select Volume Low Warn All Users and change the value to ON.

    This causes the system to send a warning message to users when a volume is becoming full. The next two steps specify when the warning is sent and when it is removed.

  5. Select Volume Low Warning Threshold and set it to the desired value.

    This parameter determines when the system will warn users that the volume is becoming full.

    The threshold specifies how much free space, in blocks, must be left on the volume for the warning to be issued.

  6. Select Volume Low Warning Reset Threshold and set it to the desired value.

    This parameter determines when the warning will be removed.

    It specifies the amount of space that must be freed on the volume for the warning message to be cleared and the warning mechanism to be reset.

    For example, both the Volume Low Warning Threshold and the Volume Low Warning Reset Threshold might be set to 256 blocks. In this case, the system sends a warning when 256 blocks of free space is left on the volume. The volume must then gain an additional 256 blocks (for a total of 512 blocks of free space) before the warning message disappears and the warning mechanism is reset.

    Once the mechanism is reset, the free space on the volume must dip below 256 blocks again before another warning is sent.

  7. Press <Esc> twice to reach the Update Options menu.

  8. Choose Update AUTOEXEC.NCF and STARTUP.NCF Now from the Update Options menu.

    A window appears, indicating the path to the AUTOEXEC.NCF file.

  9. Press <Enter> to update the file.

    If you press <Enter>, the system writes the parameters to the AUTOEXEC.NCF file or updates the parameters if they are already in the file. The file is updated immediately. You do not need to reboot the server.

    If you do not press <Enter>, the parameter changes last only until you reboot the server.


Viewing the Server Error Log

The server error log is a text file called SYS$LOG.ERR in the server's SYS_servername:SYSTEM directory. All system messages and alerts that appear on the server console are recorded in the SYS$LOG.ERR file. You should view this file periodically to see what kinds of errors are occurring on your server.

NetWare security violations are also recorded in the SYS$LOG.ERR file. Check this file daily if you are concerned about security at your site.

NOTE:  You should regularly clear the SYS$LOG.ERR file to keep it from using too much server disk space.

You can view the server error log from a workstation using a text editor or from the console usingEDIT.NLM. For more information about using EDIT, see Editing Text Files from the Server Console

You can also view or clear the server error log with the NETADMIN and FILER utilities, as described here.


Prerequisites


Procedure for NETADMIN

  1. At the workstation prompt, type

    NETADMIN <Enter>

  2. From the NetAdmin Options screen, choose Manage Objects.

  3. From the browser, select the Server object whose error file you want to view and press <F10>.

  4. Choose View or Edit Properties of This Object.

  5. Choose View Server Error Log File.

    The server error log appears.

  6. To delete or exit the server error log, press <Esc>.

    A prompt to clear the error log file appears.

  7. To clear the error log file, choose Yes.

  8. To exit NETADMIN, press <Alt>+<F10>.


Procedure for FILER

  1. At the workstation prompt, type

    FILER <Enter>

  2. From the Available Options screen, choose Manage Files and Directories.

  3. From the Directory Contents screen, browse the tree to select the SYS: volume on the Server object whose error file you want to view, and press <F10>.

  4. Choose the error log file you want to view.

  5. From the File Options screen, choose View File.

    The server error log appears.

  6. To exit the server error log, press <Esc> twice.

  7. To clear the error log file, be sure the correct filename is highlighted and press <Delete>

    A prompt to clear the error log file appears.

  8. To clear the error log file, choose Yes.

  9. To exit FILER, press <Alt>+<F10>.


Managing Error Log Files

NetWare creates and maintains three error log files:

To keep error log files from using too much disk space, NetWare 4 includes the following SET parameters:

SET Server Log File State

SET Server Log File Overflow Size

SET Volume Log File State

SET Volume Log File Overflow Size

SET Volume TTS Log File State

SET Volume TTS Log File Overflow Size

See SET in Utilities Reference for information on using these parameters to limit error log file size.


Checking for Disk Errors

You should check for disk errors regularly. Any new errors should be cause for concern, because such errors can lead to a failed file system and loss of data.

If you check for disk errors regularly, you can replace a disk before it causes data loss.

IMPORTANT:  If your server's disk drive performs read-after-write verification and automatically redirects bad blocks, the following procedure does not work. It only works if NetWare does the redirection, using Hot FixTM.


Procedure

  1. At the server console prompt, type

    LOAD MONITOR <Enter>

  2. Choose Disk Information from the Available Options menu.

  3. Choose the disk drive you want to check from the System Disk Drives menu.

    The upper part of the screen displays information about the disk device that you selected, including Hot Fix and Redirected Block statistics.

  4. Write down the number of Redirected Blocks and the number of Redirection Blocks shown in the upper window of the screen.

    When NetWare cannot reliably write a block of data to the disk, it marks the disk block as bad and writes the data to the Hot Fix Redirection Area.

    The Redirected Blocks count shown in MONITOR is the number of blocks redirected since the server was installed.

    The Redirection Blocks count is the number of blocks designated as the Hot Fix Redirection Area.

  5. Evaluate these statistics and proceed according to the following steps.

    1. Compare the current number of redirected blocks to the number you recorded the last time you checked this number.

      If the number of redirected blocks is the same as the previously noted number, you do not need to take any corrective action.

      If the number of redirected blocks has increased significantly, or if the number of redirected blocks is over half the number of redirection blocks, continue with Step 5.b.

      If the number of redirected blocks has slightly increased over the previous number, monitor the server closely.

    2. Send a message to all users to close files and log out.

    3. Back up the server data and bring down the server.

    4. Troubleshoot the disk drive, driver, controller, and host bus adapter, using the hardware documentation.


Additional Information

For more information about See

Data protection

Data protection in Concepts

MONITOR utility

MONITOR in Utilities Reference

Online help in MONITOR

Redirection, Hot Fix, and read-after-write verification

Hot Fix in Concepts