NAME

rw - User interface to RemoteWatch programs and data. 

SYNOPSIS

/usr/remwatch/bin/rw

DESCRIPTION

rw is a character-based interface for user access to RemoteWatch data and programs.  By entering choices in response to on-screen menu selections, the user can customize RemoteWatch operation, obtain online help for RemoteWatch messages, view configuration reports, and view e-mail archives.  rw can run on any node in an HP-UX cluster or on a standalone HP-UX system.  RemoteWatch is a subsystem of the HP Support Watch product. 

STARTUP

 
Any user may run rw; superuser capability is not required except for using the customize feature of rw.  The interface is accessed directly by entering: /usr/remwatch/bin/rw This brings up the RemoteWatch Main Menu. On the RemoteWatch Main Menu and on all submenus, there is a help selection.  Selecting help at the menu level brings on-screen help text about the menu.  By entering help along with a menu item number, help text for that particular item is provided.  The customization submenu also includes selections to set the parameters to defaults, to cancel any changes and exit the menu, or to save any changes made and then exit the menu.

CUSTOMIZING GENERAL PARAMETERS

The General Parameters Customization Menu displays the modifiable parameters, along with the current value of each item.  To change a value, such as the e-mail destination, select the appropriate numbered option.  The program will prompt you for the new value and perform range checking, if appropriate, on your input. Daily run
This parameter determines whether or not RemoteWatch will perform its daily run.  If enabled, RemoteWatch performs its daily run as usual.  If disabled, RemoteWatch sends a message to the configured user that it is in a disabled state and does not perform its daily run.  Disabling RemoteWatch on a cluster server will cause the daily run to be suppressed on all nodes in the cluster.  Default is "enabled". Run time (24-hour clock)
This value sets the execution time for the RemoteWatch software.  When this parameter is changed, the RemoteWatch scheduler is updated with the new run time.  If this is a clustered system, the RemoteWatch scheduler for each client is adjusted as well.  Client nodes will run at two-minute intervals after the server runs.  Default is "02:00" (2:00am). E-mail destination for messages
This value is used by RemoteWatch to mail any notification messages after each daily run.   When the user changes this value, a test message is mailed to the newly configured destination to verify that the destination is correct.  To specify more than one destination, separate the names with blanks, for example, "root jzw pilger@hpsys3", or commas, for example, "root,jzw,pilger@hpsys3".  Multiple addresses together must not exceed 80 characters. Default is "root". If you enter a destination containing the "@" character, and typing "@" causes an immediate carriage return to occur, then your system has "@" defined as the line kill character.  In this case, you should either precede the "@" with a "\" (backslash) character (e.g. "newt\@systemx") or use the stty(1) command to change the line kill character (see the man page for stty(1) for instructions). E-mail destination for alarms
This value is used by RemoteWatch to mail any alarm messages which are found throughout the day in the examination of dmesg module output.  When the user changes this value, a test message is mailed to the newly configured destination to verify that the destination is correct.  To specify more than one destination, separate the names with blanks, for example, "root fred boss@systemx", or commas, for example, "root,fred,boss@systemx".  Multiple addresses together must not exceed 80 characters.  Default is "root". If you enter a destination containing the "@" character, and typing "@" causes an immediate carriage return to occur, see the instructions above. Reports & archives expiration
RemoteWatch uses this parameter to purge old files from the reports directory and the archive directory if the specified number of days’ worth of data to be stored by RemoteWatch has been exceeded.  It is recommended that at least seven days worth of data be stored for troubleshooting purposes.  Default is "14" days. Warning threshold for reports & archives
RemoteWatch uses this parameter to monitor the amount of disk space it is allowed to use.  If the disk space used by RemoteWatch reports and archives exceeds that threshold, RemoteWatch sends a warning message to the configured user.  Default is "500" Kbytes per node. Warning threshold for system’s disks
RemoteWatch uses this parameter to monitor the amount of disk space being used on the system.  If the disk capacity of any mounted disk on the system exceeds the threshold percentage, RemoteWatch sends a warning message to the configured user.  Default is "95" percent.

LAN STATISTICS LIMITS

 
LAN Selection
On a cluster server, there is an option to change LAN limits for either one node only, or for all nodes on the cluster.  If "all nodes" is selected, one node is designated to be used as a template.  This means that this node’s limits will be used as starting values, and any changes made to these values will be applied to all nodes.  Selection of a particular node or all nodes is made by selection from an on-screen menu. Once the node or nodes selection is made, the LAN Limits Customization Menu is provided.  The 24 LAN limits, and their default values, are as follows. 
 
 1) Frames received                  - 0
 2) Frames transmitted               - 0
 3) Minimum sample size              - 10000
 4) Undelivered received frames      -  2% of frames received
 5) Untransmitted frames             -  1% of frames transmitted
 6) CRC errors received              -  2% of frames received
 7) Transmit collisions              - 10% of frames transmitted
 8) One transmit collision           -  5% of frames transmitted
 9) More transmit collisions         - 10% of frames transmitted
10) Excess retries                   -  1% of frames transmitted
11) Deferred transmissions           - 20% of frames transmitted
12) Carrier lost when transmitting   -  1% of frames transmitted
13) No heartbeat after transmission  -  1% of frames transmitted
14) Frame alignment errors           -  2% of frames received
15) Late transmit collisions         -  1% of frames transmitted
16) Frames lost                      -  5% of frames received
17) Unknown protocol                 - 10% of frames received
18) Bad control field                -  1% of frames received
19) IEEE 802.3 XID packets           -  1% of frames received
20) IEEE 802.3 TEST packets          -  1% of frames received
21) Unable to respond TEST/XID pkts  -  1% of frames received
22) Illegal sized frames             -  2% of frames received
23) Unable to find transmit buffers  -  1% of frames transmitted
24) One or zero receive buffers      -  2% of frames received
Some of these statistics are marked on the screen as being valid for the Series 700 and 800, or for the Series 800 only. NOTE: Because RemoteWatch can access RemoteWatch on remote machines running HP Support Watch for Series 800, information pertaining to Series 800 is shown here. 

Customizing LAN Limits

RemoteWatch keeps track of the statistics for each LAN card on the system. The statistics which are kept are listed above.  The purpose of tracking LAN statistics is to monitor the performance of the LAN.  Most limits (items 4 - 24) are expressed as percentages of the number of frames transmitted or received.  For example, the limit for the "CRC errors received" statistic can be set to any value between 0 and 100% of the number of frames received.  The limits for "Frames Received" and "Frames Transmitted" are expressed as absolute values and can be set to any positive integer or 0. Any absolute or percentage entry with a value of 0 causes the corresponding statistic to be ignored; no message will be sent about that statistic regardless of its value. Default values are provided for each limit, but these may need to be changed to better correspond to the level of activity on your particular LAN.  The list above shows the default values for the LAN limits.  The limits should be set to values above the usual level reached when the system and LAN are operating normally.  For example, if you consistently see a high rate of "Unknown Protocol" errors on your LAN, and you know that the LAN is operating normally, then you can use the LAN Limits Customization Menu, item #17, to change the threshold for unknown protocol errors from the default 10 percent to a higher percentage such as 15 percent or 20 percent. The "minimum sample size" (item #3) is not a LAN statistic.  For those statistics which are calculated as percentages of the number of frames transmitted or received, the LAN limits are not calculated until the configured minimum sample size (the count of frames transmitted or received) is obtained.  Setting a large sample size prevents early triggering of limits from too small a sample size.  You may wish to increase the minimum sample size if you feel that LAN limits are being calculated based on too small an amount of traffic on your LAN.

MESSAGE SUPPRESSION

RemoteWatch attempts to notify you of any situations that are identified as out of the norm for a system.  However, there may be circumstances on your system which are normal for your environment, but which RemoteWatch finds unusual.  In this case, e-mail messages would be generated daily containing information about your normal operating environment.  The message suppression facility allows you to customize RemoteWatch to your operating environment by letting you choose messages to "shut off."  The default is that message number 301 is suppressed.  This message provides dmesg(1M) command output, which many users do not wish to see in nightly e-mail messages.  There may be instances where a message is generated by RemoteWatch on a daily basis describing a condition which is completely normal for your system.  If after a time you find that this message adds no value, you may suppress that message from appearing in your nightly e-mail notification, although it will still be written to the archive file.  When messages are suppressed, the message number of those suppressed appears in the Message Suppression Customization Menu. NOTE: If you are considering suppressing LAN messages 503 or 540, please read the section about "Customizing LAN Limits" in the HP Support Watch User’s Guide.  You should raise the thresholds for the LAN statistics instead of completely suppressing these messages. 

SECURITY PARAMETERS CUSTOMIZATION

The File System Security program runs at the user-specified frequency to check file system security (refer to the HP Support Watch User’s Guide for information on the File System Security program).  The customizable parameters are the run time and run frequency for the File System Security program.  The choices for run frequency for the File System Security program are "Daily", "Weekly" (the user is prompted for the day of the week), "Monthly" (the user is prompted for the day of the month), or "Never".  The default is "04:00" (4:00am), "Weekly on Sunday".

LOG FILE MONITOR CUSTOMIZATION

The Log File Monitor program monitors user-specified files against user-specified size thresholds.  The program runs on the server or standalone system and reports on all specified log files.  It automatically checks context dependent files (CDFs).  For example, if the log file ’/etc/utmp’ is selected, all ’utmp’ files in ’/etc/utmp+’ would be monitored.  When a file size exceeds the threshold value, the configured user is notified in an e-mail message.  A default list of files to be monitored is furnished with RemoteWatch.  This file may be used as is or modified to meet the needs of the user.  The choices for Log File Monitor run time and frequency are "Daily," "Weekly," (the user is prompted for the day of the week), "Monthly" (the user is prompted for the day of the month), or "Never."  The default is daily at "04:30" (4:30 am).

CORE FILE CHECK CUSTOMIZATION

The Core File Check program runs at the user-specified frequency to check the system for core files.  Core files usually result from situations that cause a "memory fault" message from the system. If any core files are found, a message is mailed to the user.  Core files are not removed from the system by RemoteWatch.  The choices for Core File Check run time and frequency are "Daily," "Weekly," (the user is prompted for the day of the week), "Monthly" (the user is prompted for the day of the month), or "Never."  The default is weekly on Saturday at "03:30" (3:30 am).

AUTHOR

HP Support Watch and RemoteWatch were developed by the Hewlett-Packard Company. 

SEE ALSO

dmesg(1M), remotewatch(1M), stty(1), xsw(1M), HP Support Watch User’s Guide

  —  February 19, 1992