1
0
Files
irix-657m-src/eoe/cmd/sss/console/plugins/plugin_datafiles/help_sem.txt.html
2022-09-29 17:59:04 +03:00

1291 lines
63 KiB
HTML

<HTML><HEAD><TITLE>SGI Embedded Support Partner - Help</TITLE></HEAD>
<BODY BGCOLOR="#ffffcc">
<P>&nbsp;</P>
<font face="Arial,Helvetica">
<A NAME="readme_first"></A>
<b>Embedded Support Partner ASCII console usage notes</b>
<p>The SGI Embedded Support Partner ASCII console is a set of means
that provides access to the SGI Embedded Support Partner
facilities for users running cursor-addressable,
character-cell display devices (e.g., vt100 terminals,
vt100 emulators, or any other "curses-oriented" displays).
<p>In order to operate the Embedded Support Partner User Interface
from such display device, the Lynx WEB browser must be used.
It is expected that the executable file of Lynx browser will be
installed into /usr/local/bin subdirectory. Please, refer
to the Lynx's documentation about the installation of this
browser.
<p>Since there are significant differences between usage of
graphics-based Web Browser (Netscape) and ascii-based
Web Browser (Lynx) it is strongly recommended for a person
who does not have previous experience of working with Lynx
to refer to the documentation about general usage of this
WEB Browser as well as intradocument and interdocument
navigation.
<p>Due to dynamic nature of the user interface it is essential
to ensure that the HTML pages displayed by Lynx are current
and have not been loaded from Lynx's cache.
<p>To ensure that you have to follow a few simple rules:
<ol><li>Use "x" (NOCACHE on keymap) to activate links on the page
to ensure that this page is loaded from the server and not from
the cache.
<li>If you used "Backspace" or "Delete" ("HISTORY" on keymap) to
get history of the visited pages, you can must "x" (NOCACHE) to
return to the page that you selected.
<li>If you used "PREV_DOC" key to return to the previous document and
you need to refresh this page, hit "Backspace" then "x" to do the job.
</ol>
<p><b>Tip.</b> Press "k" to get current key assignment.
<A NAME="sysinfo"></A>
<hr width=100%>
<b>SYSTEM INFORMATION &gt; Introduction</b>
<P>The <b>SYSTEM INFORMATION</b> category provides information about the
system on which the Single System Manager is running.</P>
<P>Use the commands in this category to display the following types of
system information:</P>
<UL>
<LI>Hardware configuration for a specific date and time
<LI>Software configuration and version information for a specific date and time
<LI>System changes between a range of dates
<LI>Part changes for a specific hardware component
<LI>Events that have occurred on the system
<LI>Actions that the SGI Embedded Support Partner has performed
<LI>Availability information for a specified range of dates
</UL>
All reports in this category display general system information:
<UL>
<LI>System name
<LI>System identification number
<LI>System serial number
<LI>IP type
<LI>System IP address
</UL>
<A NAME="sysinfo_hardware"></A>
<hr width=100%>
<b>SYSTEM INFORMATION &gt; Hardware</b>
<P>Use this command to display the hardware configuration of the system,
which existed at a specific time on a specific date.</P>
<p>Hardware configuration information is available for the following systems:</p>
<ul>
<li>IP19 - Challenge/Onyx
<li>IP21 - Power Challenge/Power Onyx
<Li>IP25 - Power Challenge 10000/Power Onyx 10000
<li>IP27 - Origin2000/Onyx2
<li>IP29 - Origin200
<li>IP30 - Octane
<li>IP32 - O2
</ul>
<P>If you are interested in hardware information for a specific date/time, enter the desired date/time in the
appropriate field.</p>
<p>You must select a database that corresponds to the date that you specified.<p>
<P>The information is displayed in a hierarchical manner.
If information is not available or not applicable, &#034;N/A&#034; is displayed.</P>
<P>The first column of the report table can include the following symbols: &#034[+]&#034 or &#034[-]&#034.
Selecting &#034[+]&#034 symbol expands the table to display the subcomponents that compose the selected component.
Selecting &#034[-]&#034 symbol collapses the subcomponent display.</P>
The other columns of the table contain the following information:
<pre> NAME The name of the component
LOCATION The location of the component
PART_NUMBER The part number of the component
SERIAL_NUMBER The serial number of the component
REVISION The revision level of the component</pre>
<A NAME="sysinfo_software"></A>
<hr width=100%>
<b>SYSTEM INFORMATION &gt; Software</b>
<P>Use this command to display the software configuration of the system and version information
that existed at a specific time on a specific date.</P>
<p>If you are interested in software information for a specific date/time, enter the desired date/time in
the appropriate field.
Otherwise, the latest available information wil be displayed.</p>
<p>You must select the database that corresponds to the date that you specified.</p>
<p>This report lists the software that was installed on the system at the time you specify.
The installed software is listed 10 items per page.
Symbol &#034&gt;&#034 lists the next 10 pages,
&#034&gt;&gt;&#034 goes to the last page.
Symbol &#034&lt;&#034 lists the previous 10 pages, and &#034&lt;&lt;&#034 returns to the first page.
<p>The report table provides the following information:
<pre> NAME The name of the software
VERSION The version number of the software
INSTALL_DATE The date on which the software was installed
DESCRIPTION A description of the software
</pre>
<A NAME="sysinfo_system_changes"></A>
<hr width=100%>
<b>SYSTEM INFORMATION &gt; System Changes</b>
<P>Use this command to view any system changes that occurred within the range of dates that you specify.</P>
<p>If you do not specify a date, all system configuration changes are displayed.
<p>You must select the database that corresponds to the dates that you specified.</p>
<p>System change information can be collected from only one database at a time.</p>
<P>The SGI Embedded Support Partner tracks the following types of system changes:</P>
<ul>
<li>Software changes
<li>Hardware changes
<li>System changes
</ul>
<P>The software table describes all software changes that occurred during the period
of time that you specified. The table provides the following information:
<pre> NAME The name of the software
VERSION The version number of the software
INSTALL_DATE The date on which the software was installed
DEINSTALL_DATE The date on which the software was deinstalled
DESCRIPTION A description of the software
</pre>
<P>The hardware table describes all hardware changes that occurred during the period of time
that you specified. The table provides the following information:
<pre> NAME The name of the part
LOCATION The location of the part
PART_NUMBER The part number for the part
SERIAL_NUMBER The serial number of the part
REVISION The revision level of the part
INSTALL_TIME The date on which the component was installed
DEINSTALL_TIME The date on which the component was deinstalled.
</pre>
<P>The system changes table describes all system changes (for example, hostname, IP address change, and so on)
that occurred during the period of time that you specified.
<A NAME="sysinfo_part_changes"></A>
<hr width=100%>
<b>SYSTEM INFORMATION &gt; Part Changes</b>
<P>Use this command to view the transaction history of a part.</p>
<P>You must enter the component serial number. (If
necessary, use the <a href="#sysinfo_hardware">SYSTEM Information &gt;
Hardware</a> to locate a serial number.)</P>
<p>You must choose a database to view the history of
the component whose serial number you entered above.</P>
<p>The report table lists
the name of the component,
the module number in which the component was installed,
the part number of the component,
the serial number of the module,
the revision number of the part,
and the slot number in which the component was installed.</P>
<A NAME="sysinfo_events"></A>
<hr width=100%>
<b>SYSTEM INFORMATION &gt; Events Registered</b>
<P>Use this command to view information about events that SGI Embedded
Support Partner has registered.</P>
<P>Enter a range of dates for the events that you want to view.
Then, choose the type of event information that you want to view.
The following options are available:</P>
<pre>
All System Events
Specific System Event
System Events by Class
</pre>
<A NAME="screens_all_sys_events1"></A>
<p><b>All System Events</b>
<P>The report table provides the following information about events that were registered
within the selected range of dates:</P>
<pre>
Event Class The class in which the event belongs
(for example, Availability)
Event Description A brief description of the event
Event ID The unique identification number assigned
to this event. You can use this number to
find this event via SYSLOG
First Occurrence The date and time that the event first occurred
Last Occurrence The date and time that the event last occurred.
If Number of Occurrences is 1, the time value of
the First Occurrence and the time value of the
Last Occurrence will be identical
Number of Occurrences The number of times that the event occurred.
This number corresponds to the number of events
that must occur before registration begins.
By default, this number is 1.
</pre>
<A NAME="screens_specific_sys_event1"></A>
<p><b>Specific System Event</b>
<P>Use this report to track a specific event that is associated with an actual or suspected
system problem. Choose an event class from the list that appears.</P>
<A NAME="screens_specific_sys_event2"></A>
<P>Use this page to specify the event that you want to view. Choose the
event from the list of events in the class that you have already specified.</P>
<A NAME="screens_specific_sys_event3"></A>
<P>The report table provides the following information about the event registrations between the
selected range of dates:</P>
<pre>
First Occurrence The date and time that the event first occurred
Last Occurrence The date and time that the event last occurred.
If Number of Occurrences is 1, the time value of
the First Occurrence and thetime value of the
Last Occurrence will be identical.
Number of Events The number of times that the event occurred.
This number corresponds to the number of events
that must occur before registration begins.
By default, this number is 1.</TD>
</pre>
<A NAME="screens_sys_event_class"></A>
<P><b>System Events by Class</b>
<P>Use this report when you need information about events that are associated with a specific
class. For example, use Memory class to track various memory
events. Choose the appropriate class for the event that you want to view. </P>
<A NAME="screens_sys_event_class2"></A>
<P>The report table provides the following information about events that were registered between the
selected range of dates: </P>
<pre>
Event Description A brief description of the event
Event ID The unique identification number
assigned to this event
First Event Occurrence The date and time that the event first occurred
Last Event Occurrence The date and time that the event last occurred.
If Number of Occurrences is 1, the time value of
the First Occurrence and the time value of the
Last Occurrence will be identical
Number of Events The number of times that the event occurred.
This number corresponds to the number of events
that must occur before registration begins.
By default, this number is 1.
</pre>
<A NAME="sysinfo_actions"></A>
<hr width=100%>
<b>SYSTEM INFORMATION &gt; Actions Taken</b>
<P>Use this command to display information about actions that have been performed
by SGI Embedded Support Partner.</P>
<p>Specify the range of dates for which you want to report actions taken.
If you do not enter a date, this option defaults to the current date.</p>
<p>You must choose one of the two available types of reports:</p>
<pre>
All Actions Taken
Actions Taken for a Specific Event
</pre>
<a name="all_actions_taken"></a>
<p><b>All Actions Taken</b></p>
<P>This option displays the actions that the SGI Embedded
Support Partner performed within the range of dates that you specified.
The report table provides the following information about actions that were taken for all events
between the selected range of dates:</P>
<pre>
Event Class The class in which the event belongs
for example, Availability)
Event Description A brief description of the event
Event ID The unique identification number
assigned to this event
Action Description A brief description of the action
Action Taken The action that SGI Embedded Support Partner
performed in response to the event
Time of Action The date and time that SGI Embedded Support Partner
performed the action
</pre>
<a name="specific_action_taken"></a>
<p><b>Actions Taken for a Specific Event</b></p>
<p>Use this option when you want to view actions taken for specific events.
Choose an event class that contains the event that you want to select.</p>
<a name="specific_action_taken1"></a>
<p>From the list of events, choose the event that you want to research.</p>
<P>The report table provides the following information about actions that were taken for the
specified event between the selected range of dates:</P>
<pre>
Action Description A brief description of the action
Action Taken The action that the SGI Embedded Support Partner
performed in response to the event
Time of Action The date and time that SGI Embedded Support Partner
performed the action
</pre>
<a name="report_diags_result"></a>
<p><hr>
<b>Diagnostics Results</b>
<p>This command displays the results of the diagnostics that you run on the system.
<p>You must specify the range of dates for which you want to view diagnostics results.
<p>The top portion of the diagnostic report contains the information that pertains
to the system from which you requested the report.
<p>The diagnostics results table provides the following information for all diagnostics that were run
on the system during the period of time that you specified:
<p>
<pre>
Diagnostic Name Contains the name of diagnostic.
In cases where multiple tests run as a group under
one program (for example, under SVP), the total
number of tests is indicated in parentheses next
to the name of the diagnostic:
SVP (86) means that 86 tests ran under
the SVP program.
Diagnostic Status Diagnostic status can be PASS, FAIL or COMPLETE.
PASS indicates that the diagnostic completed
successfully
FAIL indicates a failure occurred
COMPLETE indicates that multiple tests ran, and
one or more of them failed and others
completed successfully
Diagnostic Result
Time The time when the diagnostic test completed.
When multiple tests run under one program, the
Diagnostic Result Time indicates the time when
the entire program completed.
</pre>
<p>
<A NAME="sysinfo_availability"></A>
<hr width=100%>
<b>SYSTEM INFORMATION &gt; Availability</b>
<P>This command displays system availability statistics. The upper portion of this page displays the total availability percentage
and the mean time between interrupts (MTBI) in minutes.</P>
<P>You must specify the range of dates and type of availability information that you want to view.
Two types of availability information are curently available:
<pre>
Overall Availability
Availability Events List
</pre>
<A NAME="screen_over_avail"></a>
<p><b>Overall Availability</b>
<p>The <b>Overall Availability</b> covers the aggregation of events for the given system.
Events are grouped as either "Unscheduled" or "Service Action" (controlled shutdown) events. Events are
further classified by categories within these two groups. For each category, overall availability report includes
the count of events in that category, the total downtime (in minutes), the MTBI (mean time between interrupts, in
minutes) and the availability as a percentage. MTBI and availability per category are computed for events within the
category as applied to the entire time period of the report. Count, total downtime, MTBI, and availability are also
displayed for the two groups, as well as the final total of all the events.</p>
<p>The average, least, and most uptimes and downtimes are also included in the report in addition to logging start time
and the duration of system uptime since the last boot.</p>
<P>The <b>Overall Availability</b> table summarizes the overall availability of the system:</P>
<UL>
<LI>Information about service actions (number, downtime, MTBI, and availability percentage)
<LI>Average uptime
<LI>Least uptime
<LI>Most uptime
<LI>Average downtime
<LI>Least downtime
<LI>Most downtime
<LI>The time at which availability monitoring was started
<LI>The time the last boot occurred
<LI>The amount of time that the system has been up
</UL>
<P>Use the <b>Event Availability Information</b> link at the bottom of the page
to access information about the individual availability events that the system has registered.</P>
<p><b>Event Availability Information</b></p>
<p>In the events list display, the fields shown are Start time (when the system was previously booted), the Incident Time,
when the event occurred, the uptime and downtime in minutes, and a very brief description of the event type or cause of the
event. The <b>Summary</b> displays the event information with more details, including a complete event type description.</p>
<a name="event_summary"></a>
<P>The report provides a summary of an event that includes the following information:</P>
<UL>
<LI>The hostname of the system
<LI>The reason for the shutdown
<LI>The time that the system was initially started
<LI>The time that the incident occurred
<LI>The time that the system was restarted after the incident occurred
<LI>The amount of time that the system was up before the incident occurred
<LI>The amount of time that the system was down because of the incident
</UL>
<p>If a system panic occurs, this report also includes a brief summary of why the system panicked.</p>
<A NAME="setup_intro"></A>
<hr width=100%>
<p><b>Setup &gt; Introduction</b></p>
<p>Embedded Support Partner is a configuration driven system.&nbsp; From this
section, you can setup SGI Embedded Support Partner to suit your specific
needs. On the left is a menu consisting of various items organized
in groups each of which belongs to a specific component of SGI Embedded
Support Partner. A brief description of these components is given below.
A context sensitive help is also available for all applicable menu items
and can be viewed by selecting 'Help' button on the top right-hand corner
of the menu item. You can always view the current settings by selecting
'View Current Setup' item for any of the components.<br>&nbsp;
<ul>
<li>
<b>Global Setup</b> allows you to setup permissions for other systems to connect
remotely to SGI&nbsp;Embedded Support Partner.&nbsp; This can be accomplished
by selecting Server item.&nbsp; Global Configuration lets you&nbsp; modify
the behavior of SGI&nbsp;Embedded Support Partner Event Manager.<br>&nbsp;</li>
<li>
<b>Events Setup</b> lets you add new events, update existing events and delete
any unwanted events.&nbsp; You can also associate an action to an event
from a list of available actions.&nbsp; Please choose Actions setup to
add custom actions.<br>&nbsp;</li>
<li>
In <b>Actions Setup</b>, you can create new actions or update existing actions
or delete any unwanted actions.<br>&nbsp;</li>
<li>
The <a href="#setup_notification">Paging Setup</a> lets you configure QuickPage application to suit your
needs.&nbsp; Please note that in order for Paging Setup to take effect,
you must chkconfig quickpage on.<br>&nbsp;</li>
<li>
<a href="#setup_availmon">Availability Monitoring</a> Setup lets you configure availmon that is available
on your system.&nbsp; You can set the parameters of Availmon and you can
also setup the mailing lists to whom availmon will notify any system interrupts.<br>&nbsp;</li>
<li>
<b>Performance Monitoring</b> Setup lets you enable or disable performance monitoring
metrics.<br>&nbsp;</li>
</ul>
<p>Caution must be observed while changing any of the settings.&nbsp; If you
are in doubt, please read the help carefully before committing any changes.&nbsp;
You can also refer to SGI&nbsp;Embedded Support Partner User Guide for
more information.
<A NAME="setup_global_web_access_cfg"></A>
<p><b>SETUP &gt; Global &gt; Server</b>
<p>This command configures the Web server that SGI Embedded Support Partner uses.
Use this command to perform the following functions:</P>
<UL>
<LI>Display the current server port, version number, and identification information
<LI>Specify access privileges to the system via IP addresses
<LI>Change the username of the current Web server user
<LI>Change the password of the current Web server user
</UL>
<P>The upper portion of this page displays the following information:</P>
<pre>
Server Identification The name of the Web server software in use
Server Version The version level of the Web server
software and its installation date
Server Port The Web server connection port in use
</pre>
<P>The lower portion of this page displays the following selectable options:</P>
<pre>
Server Access Permissions Enables or restricts access
by external systems
Name &amp; Password Change Enables you to change the current
username and password
</pre>
<A NAME="server_access_option"></A>
<p><b>Server Access Permissions</b>
<P>Use this page to specify which systems can access the SGI Embedded
Support Partner Web server. Any change that you make to the server access
list takes effect immediately.</P>
<p>You can specify the exact IP address or IP address mask using a wildcard. For example,
197.23.14.5, or 135.*.*.5, or *.*.*.*, and so on.</p>
<A NAME="user_name_change_option"></A>
<b>User Name and Password Change</b>
<P>Use this page to change a current username or password that enables access to
SGI Embedded Support Partner. Any change that you make to a username or
password takes effect immediately.</P>
<p>The username and password must each contain between 1 and 128 characters.
Characters like &#034;*&#034;, &#034;&amp;&#034;, and &#034;:&#034; are not allowed in the username
and password strings.</p>
<p>The default username <b>administrator</b> and the default password
<b>partner</b> must be changed immediately after installation.</p>
<A NAME="setup_global_event_cfg"></A>
<hr width=100%>
<b>SETUP &gt; Global &gt; Global Configuration</b>
<P>An <I>event</I> is a happening or an occurrence that takes place on the
system that SGI Embedded Support Partner is monitoring. A few examples of
events follow: parity errors, disk full, nonmaskable interrupts (NMI), and
even activities of the SGI Embedded Support Partner itself.</P>
<P>Use this page if you want to reset the following parameters for all events on the system.
<UL>
<LI>The <b>Log events</b> parameter enables or disables global
event logging. Select <b>Log events</b> checkbox to log events in the SGI
Embedded Support Partner database. Deselect <b>Log events</b> if you
do not want to log events in the SGI Embedded Support Partner database.
You can disable event logging if you are not interested in the history of
events on the system.<p>
<LI>The <b>Throttle events</b> parameter enables or disables event
registration requirements for all events. Select <b>Throttle events</b> checkbox to require that a
specific number of events (a threshold) must occur before the event is registered in the
SGI Embedded Support Partner database. Deselect <b>Throttle events</b> checkbox to register
every event in the SGI Embedded System Partner database. Enable event throttling,
if you are not interested in every event of a particular type, but you are interested only
when this event occurs a specified number of times.<p>
<LI>The <b>Act on event</b> parameter enables or disables the capability of SGI Embedded Support
Partner to react (respond) to events. Select <b>Act on event</b> checkbox to specify that the
SGI Embedded Support Partner should respond (react) to all events.
Deselect <b>Act on event</b> checkbox to specify that the SGI Embedded Support Partner should
not respond (react) to any events.
<p>Note: Refer to the SETUP &gt; Events and the SETUP &gt; Actions menus for
additional information about events and actions.
</UL>
<p>Note: The Global Configuration setting will override individual event setting.</p>
<A NAME="setup_events_viewcurr"></A>
<hr width=100%>
<b>SETUP &gt; Events &gt; View Current Setup</b>
<P>Because the number of events can be extensive, events are divided into sets called <I>classes</I>. This scheme simplifies
the management of events, enables more efficient use of displays, and facilitates navigation within
the program.</P>
<p>The following options are available:</p>
<ul>
<li>View Event
<li>View Event List
<li>View Classes
</ul>
<a name="setup_view_event"></a>
<p><b>View Event</b>
<p>Use this option to determine the current setting of an individual event. This option allows you to view:
<ul>
<li>Event class ID and class name
<li>Event ID and event description
<li>Event registration
<Li>Number of events that must occur per registration
<li>Actions for the specified event
</ul>
<a name="setup_view_event_list"></a>
<p><b>View Event List</b></p>
<p>Use this option when you want to obtain a list of all events compatible with the SGI Embedded Support Partner.
The report allows you to view:</p>
<ul>
<li>Class Name
<li>Event Description
</ul>
<a name="setup_view_classes"></a>
<p><b>View Classes</b></p>
<p>Use this option when you want to view all classes available on the system. The report allows you to view:</p>
<ul>
<LI>Class ID
<LI>Class Name
</ul>
<A NAME="screens_setup_events_update"></A>
<hr width=100%>
<b>SETUP &gt; Events &gt; Update</b>
<p>Use this command to update (change settings for ) an existing event.
Only one event at a time can be updated using Ascii console.
<A NAME="screens_setup_events_update_settings"></A>
<p><b>SETUP &gt; Events &gt; Update &gt; Change Settings</b>
<p>1. Set checkmark to enable the registration of chosen event with SGI Embedded Support Partner.
Remove checkmark to disable the registration of chosen event with SGI Embedded Support Partner.
<p>2. Enter the number of events that must occur before registration begins.
<p>3. Select <b>Accept</b> button to set your changes.
<p>4. Select <b>Change Action Settings</b> link to change the action(s) that will be taken upon the
occurrence of the chosen event.
<p>5. Select <b>Return to Update &gt; Select Event page</b> link to select another event.
<A NAME="setup_events_update_event_actions"></A>
<p><b>SETUP &gt; Events &gt; Update Actions</b></p>
<P>An <b>event/action assignment</b> defines the action that the SGI Embedded
Support Partner performs when it registers a specific event. An <b>event/action</b>
is a cause-and-effect relationship between an event and an ensuing action.
Use this command to modify an event/action assignment; that is, to replace,
add, or delete event/action assignments.</P>
<p>In order to Update event/action relationship you must:
<p>1. Select the event for which you want to update the action assignment.</P>
<p>2. Select <b>Change Action Settings</b> link on <b>SETUP &gt; Events &gt; Update &gt; Change Settings</b> page.
The list of actions that are curently available will be displayed.</p>
<p>3. Select actions that you want to be assign to chosen event.</p>
<p>4. Select <b>Accept</b> button to assign selected actions.</p>
<p>5. Select <b>Return to Update Event page</b> link to return to <b>SETUP &gt; Events &gt; Update &gt; Change Settings</b> page</p>
<A NAME="setup_events_add"></A>
<hr width=100%>
<b>SETUP &gt; Events &gt; Add</b>
<P>Use this command to add new events for the SGI Embedded Support
Partner to monitor.
<pre>
To add the new event you must:
1. Using provided listbox, specifies the existing class
to which you want to add the new event
OR
Set checkmark, if you want to create the new class for this event,
and enter a new class name in the next input field.
<b>Note</b>. The checkmark must be removed in order to add
the new event into an existing class.
2. Enter a name for the new event
3. Specifies a description of the event that is shown in the interface
4. Set checkmark to enable the registration of this event with
SGI Embedded Support Partner
5. Enter the number of events that must occur before registration begins
6. Press <b>Accept</b> button to add the new event
OR
Press <b>Clear</b> button to clear fields and start from the beginning.
</pre>
<a name="setup_events_delete"></a>
<hr width=100%>
<p><b>SETUP &gt; Events &gt; Delete Custom Events</b></p>
<p>Use this command to delete custom event(s) from the SGI Embedded Support Partner.
All records and information associated with these classes/events will also be deleted.
Empty classes will be automatically deleted.
<pre>
In order to select event(s) to be deleted you must
Press <b>Show all custom events</b> button
to display the list of all custom events
OR
Choose the event class and
press <b>Show custom events for selected class</b> button
to display the list of all custom events for selected class.
</pre>
<a name="setup_events_delete1"></a>
<pre>
Set checkmarks for the event(s) that you want to delete.
Press <b>Delete Selected Events</b> button.
</pre>
<A NAME="setup_actions_viewcurr"></A>
<hr width=100%>
<b>SETUP &gt; Actions &gt; View Current Setup</b>
<P>Use this command to view the current configuration of actions. The following options are available:</P>
<pre>
View Action Setup Displays the configuration information
for a specific action
View Available Actions List Displays a table of all actions
that are currently available
</pre>
<A NAME="screens_action_setup"></A>
<p><b>View Action Setup</b></p>
<p>You must choose an action whose information you want to view.</p>
<p>This option allows you to view the following action information:</p>
<ul>
<li>Action command string - an exact command that will execute
<li>Action description - simple description of the action
<Li>Execute this action as nobody, guest, etc. (default = nobody)
<li>Action timeout (default = 600 seconds)
<li>Number of times the event must be registered before an action will be taken (default = 1)
<li>Retry times (default = 0)
</ul>
<A NAME="screens_avail_action"></A>
<p><b>View Available Actions List</b></p>
<P>This report displays all actions that are currently available.
The table includes the following information:</P>
<ul>
<li>Action order number
<li>Action command string - an exact command that will execute
<li>Action description - simple description of the action
</ul>
<A NAME="setup_actions_update"></A>
<hr width=100%>
<b>SETUP &gt; Actions &gt; Update</b>
<P>Use this command to update an existing action.</p>
<p>Select an action that you want to update. You can modify all of the action parameters, except the action description:</p>
<pre>
Actual action command string Specifies the command that action executes
A username to execute the action Specifies the user account that the SGI
as (Default = nobody) Embedded Support uses to execute the
command
Action timeout Specifies the time period for which the
action can run without being killed.
The value that you specify must be a
multiple of 5. (Default = 600 seconds)
The number of times that Specifies how many times the event must be
the event must be registered registered before the SGI Embedded Support
before an action will be taken Partner performs this action
The number of retry times Specifies the number of times that the SGI
Embedded Support Partner attempts to
execute the action before it stops.
The value cannot exceed 23; however, it is
not recommended to set it greater than 4.
</pre>
<a name="setup_action_example"></a>
<pre>
For example:
action to run is <b>diagnostic</b>
username to execute an action <b>nobody</b>
action timeout <b>3600</b>
the number of times that <b>5</b>
the event must be registered
before an action will be taken
the number of retry times <b>2</b>
</pre>
<p>This diagnostic will run after the event is registered in the SGI Embedded Support Partner database 5 times.
It will be executed with nobody privileges. If the diagnostic is still running after an hour (3600 seconds),
it will be killed and restarted a second time (retry times = 2).
<A NAME="setup_actions_add"></A>
<hr width=100%>
<b>SETUP &gt; Actions &gt; Add</b>
<P>Use this command to add a new action. The following options are available:</P>
<pre>
Action description Provides a description of the action.
Example: page to John Dow
Action command string Specifies the exact action command
to execute.
Example: /usr/bin/espnotify -p 1234567
Username to execute the action Specifies the user account that the SGI
as (default = nobody) Embedded Support uses to execute the
command. (Default = nobody)
Action timeout Specifies the time period for which the
action can run without being killed.
The value that you specify must be a
multiple of 5. (Default = 600 seconds)
The number of times an event Specifies how many times the event must
must be registered before an be registered before the SGI Embedded
action will be taken Support Partner performs this action.
The number of retry times Specifies the number of times that
the SGI Embedded Support Partner attempts
to execute the action before it stops.
The value cannot exceed 23; however,
it is not recommended that you set it
greater than 4.
</pre>
<pre>
For example:
action to run is <b>diagnostic</b>
username to execute an action <b>nobody</b>
action timeout <b>3600</b>
the number of times that <b>5</b>
the event must be registered
before an action will be taken
the number of retry times <b>2</b>
</pre>
<p>This diagnostic will run after the event is registered in the SGI Embedded Support Partner database 5 times.
It will be executed with nobody privileges. If the diagnostic is still running after an hour (3600 seconds),
it will be killed and restarted a second time (retry times = 2).
<P>Examples of notification options:</p>
<ul>
<li>/usr/bin/espnotify -E your_email@sgi.com,your_email2@sgi.com -e email_subject (email notification)
<li>/usr/bin/espnotify -p pager_id (pager notification)
<li>/usr/bin/espnotify -A message_string (display message on the console)
<li>/usr/bin/espnotify -D your_system_name:0.0 -c %D (graphical pop-up window with an event data in it)
</ul>
<p>For more information regarding notification options, refer to the <b>espnotify</b> man page.</p>
<p>The following list includes the accepted user format strings and any action-specific options:</p>
<ul>
<li>%C = event class
<li>%T = event type
<li>%D = event data (this is the data received along with the event.)
<li>%H = host name from which event originated
<li>%S = Event time stamp, time event occured (in seconds since Jan 1 1970)
<li>%F = forwarder hostname (in case of DSM.)
<li>%I = sys id
<li>%t = time string (current)
<li>%s = seconds since Jan 1 1970 (current)
<li>%m = current minute of the hour 0-59 (current)
<li>%M = current month of the year 0-11 (current)
<li>%h = current hour of the day 0-23 (current)
<li>%y = current year (current)
<li>%d = day of the month (current)
</ul>
<p>For example: /usr/bin/espnotify -D system_name.sgi.com:0.0 -c %D</p>
<p>This displays a window on the machine system_name.sgi.com. The window contains data that is significant to the event.</p>
<A NAME="setup_actions_delete"></A>
<hr width=100%>
<b>SETUP &gt; Actions &gt; Delete</b>
<P>Use this command to delete an action. Choose an action that you want to delete.</p>
<p>Note: The action will be deleted from the SGI Embedded Support Partner database. If this action is
assigned to some events, the list of all affected events is displayed. You have a choice to cancel or proceed with deletion.
Press <b>Yes</b> button to delete the action and remove the selected action from all events to which
it is assigned. To cancel operation return back to previous page.
<A NAME="setup_notification"></a>
<hr width=100%>
<b>SETUP &gt; Paging</b>
<p>Use <b>espnotify</b> action to deliver a text/numeric message to a pager by specifying appropriate
command line options. You may obtain more information on espnotify by using the <b>man espnotify</b> command.</p>
<p>To work properly, paging has to be configured. The SGI Embedded Support Partner provides the User Interface to set required
configuration parameters. All the parameters are written to <b>/etc/qpage.cf</b> file.</p>
<p><b>Paging</b> requires that a modem be connected to the system to dial the paging service provider
to deliver a page. The Modem/Admin section enables modem configuration. The Service section enables configuration of the parameters
of the Paging Service Provider(s). Because the service provider normally identifies each individual pager by means of a pager ID
(which does not have to be the pager Touch-tone number), a pager ID must be provided in order to deliver the page. The Pager section
enables you to configure different pagers that are associated with the Service.</p>
<A NAME="setup_notification_viewcurr"></A>
<hr width=100%>
<b>SETUP &gt; Paging &gt; View Current Setup</b>
<P>Use this command to display the current values of the paging parameters
and the following types of information:</P>
<UL>
<LI><b>espnotify</b> Administration Variables
<LI>Modem Setup Parameters
<LI>Services Setup Parameters
<LI>Pager Setup Parameters
</UL>
<A NAME="setup_notification_modem"></A>
<hr width=100%>
<b>SETUP &gt; Paging &gt; Modem/admin</b>
<P>You can configure the following <b>Modem setup</b> parameters:</P>
<p><b>Modem name</b><br>
Specifies a unique name that the SGI Embedded
Support Partner uses to identify a modem. Entering an existing modem name will update the modem name.
No spaces are allowed.
<p><b>Modem device</b><br>
Specifies the device to which the modem is connected (for example, <b>/dev/ttya</b>)
<p><b>Modem initialization command</b><br>
Specifies the command that the SGI Embedded Support Partner should use to initialize the modem
before dialing the Service Provider. These initialization commands are modem specific and are available
in your modem manual. For example, many paging services require that error correction be turned off on your modem.
For some modems, this can be done by including &amp;A0&amp;K0&amp;M0 in the modem initialization command
<P>You can configure the following <b>Administration Setup</b> parameters:</P>
<pre>
Administrator's e-mail address Specifies the e-mail address of
the person to contact if <b>Paging</b>
fails to deliver a pager
The time interval for retrying Specifies the amount of time that the
<b>espnotify</b> should wait between retries
</pre>
<A NAME="setup_notification_service"></A>
<hr width=100%>
<b>SETUP &gt; Paging &gt; Service</b>
<P>Use this command to set up information about a paging service.</P>
<P>You can configure the following parameters:</P>
<p><b>Service name</b><br>
Specifies the unique name that the SGI Embedded Support Partner uses to identify paging service provider.
Entering an existing service name will result update the service name. No spaces are allowed.
<p><b>Device</b><br>
Specifies the device (modem name) that the SGI Embedded Support Partner should use to dial the service provider.
Use <a href="#setup_notification_modem">SETUP &gt; Paging &gt; Modem/Admin</a> to set up any modems.
<p><b>Maximum number of retries</b><br>
Specifies the maximum number of times the SGI Embedded Support Partner should attempt to access this service
before it quits trying.
<p><b>Maximum length of the message</b><br>
Specifies the maximum number of characters that can be sent using this service. This depends on your service provider.
<p><b>Phone number of the paging service</b><br>
Specifies the IXO/TAP telephone number of the Service Provider. Do not confuse your pager's Touch-tone telephone
number with the service provider's IXO/TAP telephone number. They are never the same.<p>The telephone number
should contain at least 7 numbers and should not include any spaces, &#034;-&#034;, or other symbols.
<A NAME="setup_notification_pager"></A>
<hr width=100%>
<b>SETUP &gt; Paging &gt; Pager</b>
<P>Use this command to set up a specific pager.</P>
<P>You can configure the following parameters:</P>
<pre>
Pager Name Specifies a unique name to identify this pager
Pager ID Specifies the ID that is used by your Paging
service provider to identify the pager.
The ID is not necessarily be the touch-tone
phone number that you dial to access the pager.
Please, contact your service provider to get
this information.
Service Name Specifies the paging service (service name) to which
<b>espnotify</b> should deliver the page for this pager
Use the <a href="#setup_notification_service">SETUP &gt; Paging &gt; Service</a>
to set up any paging services that you want to use</a>
</pre>
<A NAME="setup_availmon"></a>
<hr width=100%>
<b>SETUP &gt; Availability Monitoring</b>
<p>The Availability Monitoring is a set of tools that collectivly monitor and report the availability of
system(s) and diagnosis of system crashes. Availability monitoring tools gather information from diagnostic
programs like ICRASH, FRU Analyzer, SYSLOG and identify the cause of system shutdowns. The system configuration
information comes from configmon, hinv and versions. Availability monitoring tools can report data to various locations
based on the Availability MailList setting.
<A NAME="setup_availmon_viewcurr"></a>
<hr width=100%>
<b>SETUP &gt; Availability Monitoring &gt; View Current Setup</b>
<P>Use this command to view the current values of the availability monitor parameters. It displays the
following information:</P>
<UL>
<LI>General availability monitoring parameters
<LI>Availability monitoring e-mail list parameters
</UL>
<A NAME="setup_availmon_configuration"></a>
<hr width=100%>
<b>SETUP &gt; Availability Monitoring &gt; Configuration</b>
<P>Use this command to set up the <font face="Courier"><TT>availability monitor</TT></font> component of
the SGI Embedded Support Partner.</P>
<P>You can configure the following parameters:</P>
<pre>
<b>Automatic e-mail distribution</b> (Enable or Disable)
Specifies whether <b>availability monitor</b> should
automatically distribute reports by e-mail.
<b>Display of shutdown reason</b> (Enable or Disable)
Specifies whether <b>availability monitor</b> should
display the reason for a shutdown
<b>Include HINV information into e-mail</b> (Yes or No)
Specifies whether <b>availability monitor</b> should
include HINV information in the diagnostic e-mail
messages that it generates.
<b>Capturing of important system messages</b> (Enable or Disable)
Specifies whether <b>availability monitor</b> should
capture important system messages.
<b>Start uptime daemon</b> (Yes or No)
Specifies whether <b>availability monitor</b> should
start the uptime daemon
<b>Number of days between status updates</b> (0 - 300,Default-60)
<b>Availability monitor</b>, using eventmond, periodically sends a status
report if the system is up for an extended period of time. This value
specifies the number of days after which a status report should be sent.
<b>Interval in seconds between uptime check </b> (User specified)
Specifies the number of seconds that <b>event monitor</b>
should wait before it performs an uptime check on the system.
(default = 300 seconds)
</pre>
<A NAME="setup_availmon_email"></a>
<hr width=100%>
<b>SETUP &gt; Availability Monitor &gt; e-mail List</b>
<P>Use this command to set up the e-mail lists for availability information reports.</P>
<P>You can set up e-mail lists for the following reports:</P>
<UL>
<LI>Availability report in text format
<LI>Availability report in compressed format
<LI>Availability report in compressed format (encrypted)
<LI>Diagnostic report in text format
<LI>Diagnostic report in compressed format
<LI>Diagnostic report in compressed format (encrypted)
<li>Pager report in concise text form
</UL>
<P>The availability report contains computed system availability metrics.</P>
<P>The diagnostic report includes all of the availability report data and
diagnostic data for troubleshooting.</P>
<a name="setup_performance_viewcurr"></a>
<hr width=100%>
<b>SETUP &gt; Performance Monitoring &gt; View Current Setup</b>
<p>All performance rules can be enabled or disabled via user interface.Use this command to display performance rules status.</p>
<p>The report table displays the following information:</p>
<ul>
<li>PMIE Rule Description
<li>PMIE Rule
<li>Status (enabled/disabled)
</ul>
<a name="setup_performance_config"></a>
<hr width=100%>
<b>SETUP &gt; Performance Monitoring &gt; Configuration</b>
<p>There is a set of rules available to set up for performance monitoring.</p>
<p>The table below provides a short description for each rule:</p>
<p><pre> <b>cpu.context_switch</b> High aggregate context switch rate</pre>
Average number of context switches per CPU per second exceeded threshold over the past sample interval.
<p><pre> <b>cpu.excess_fpe</b> Possible high floating point exception rate</pre>
This predicate attempts to detect processes generating very large
numbers of floating point exceptions (FPEs). Characteristic of
this situation is heavy system time coupled with low system call
rates (exceptions are delivered through the kernel to the process,
taking some system time, but no system call is serviced on the
application's behalf).
<p><pre> <b>cpu.load_average</b> High 1-minute load average</pre>
The current 1-minute load average is higher than the larger of
min_load and ( per_cpu_load times the number of CPUs ).
The load average measures the number of processes that are running,
runnable or soon to be runnable (i.e. in short term sleep).
<p><pre> <b>cpu.low_util</b> Low average processor utilization</pre>
The average processor utilization over all CPUs was below threshold
percent during the last sample interval.
This rule is effectively the opposite of cpu.util and is disabled by
default - it is only useful in specialized environments where, for
example, processing is batch oriented and low processor utilization
is indicative of poor use of system resources. In such a situation
the cpu.low_util rule should be enabled, and cpu.util disabled.
<p><pre> <b>cpu.syscall</b> High aggregate system call rate</pre>
Average number of system calls per CPU per second exceeded
threshold over the past sample interval.
<p><pre> <b>cpu.system</b> Busy executing in system mode</pre>
Over the last sample interval, the average utilization per CPU was
busy percent or more, and the ratio of system time to busy time
exceeded threshold percent.
<p><pre> <b>cpu.util</b> High average processor utilization</pre>
The average processor utilization over all CPUs exceeded threshold
percent during the last sample interval.
<p><pre> <b>craylink.node_cb_errs</b> CrayLink checkbit errors on Origin node</pre>
For some Origin 2000 node, at least one checkbit error was
observed on the node (CrayLink) interface and/or the I/O interface in the last sample interval. Use the command
<br><center>$ pminfo -f hinv.map.node</center> to discover the abbreviated PCP names of the installed nodes and
their corresponding full names in the <b>/hw</b> file system.
<p><pre> <b>craylink.router_cb_errs</b> CrayLink checkbit errors on Origin route</pre>
For some CrayLink router port, at least one checkbit error was
observed in the last sample interval. Use the command
<br><center>$ pminfo -f hinv.map.routerport</center>
to discover the abbreviated PCP names of the installed router ports
and their corresponding full names in the <b>/hw</b> file system.
<p><pre> <b>filesys.buffer_cache</b> Low buffer cache read hit ratio</pre>
Some filesystem read activity (at least min_lread Kbytes per
second of logical reads), and the read hit ratio in the buffer
cache is below threshold percent.Note: It is possible for the read hit ratio to be negative
more phsical reads than logical reads) - this can be as a result of:
<ul><li>XLV striped volumes, where blocks span stripe boundaries
<li>very large files, where the disk controller has to read
blocks indirectly (multiple block reads to find a single
data block result)
<li>file system read-ahead pre-fetching blocks which are not
subsequently read
</ul>
<p><pre> <b>filesys.dnlc_miss</b> High directory name cache miss rate</pre>
With at least min_lookup directory name cache (DNLC) lookups per
second being performed, threshold percent of lookups result in
cache misses.
<p><pre> <b>filesys.filling</b> File system is filling up</pre>
Filesystem is at least threshold percent full and the used space
is growing at a rate that would see the file system full within
lead_time.
<p><pre> <b>memory.exhausted</b> Severe demand for real memory</pre>
The system is swapping modified pages out of main memory to the
swap partitions, and has been doing this at the rate of at least
threshold pages swapped out per second for at least pct of the last
10 samples, ie. sustained page out activity.
<p><pre> <b>memory.swap_low</b> Low free swap space</pre>
There is only threshold percent swap space remaining - the system
may soon run out of virtual memory. Reduce the number and size of
the running programs or add more swap(1) space before it completely
runs out.
<p><pre> <b>network.buffers</b> Serious demand for network buffers</pre>
During the last sample interval the rate at which processes tried to
acquire network buffers (mbufs) and either failed or were stalled
waiting for a buffer to be freed is greater than threshold times per
second.
<p><pre> <b>network.tcp_drop_connects</b> High ratio of TCP connections dropped</pre>
There is some TCP connection activity (at least min_close
connections closed per minute) and the ratio of TCP dropped
connections to all closed connections exceeds threshold percent
during the last sample interval. High drop rates indicate either
network congestion (check the packet retransmission rate) or an
application like a Web browser that is prone to terminating TCP
connections prematurely, perhaps due to sluggish response or user
impatience.
<p><pre> <b>network.tcp_retransmit</b> High number of TCP packet retransmissions</pre>
There is some network output activity (at least 100 TCP packets per
second) and the average ratio of retransmitted TCP packets to output
TCP packets exceeds threshold percent during the last sample
interval. High retransmission rates are suggestive of network congestion, or
long latency between the end-points of the TCP connections.
<p><pre> <b>per_cpu.context_switch</b> High per CPU context switch rate</pre>
The number of context switches per second for at least one CPU
exceeded threshold over the past sample interval. This rule only applies to multi-processor systems, for
single-processor systems refer to the cpu.context_switch rule. For Origin 200 and Origin 2000 systems, use the command
<br><center>$ pminfo -f hinv.map.cpu</center> to discover the abbreviated PCP names of the installed CPUs and
their corresponding full names in the <font face="Courier"><tt>/hw</tt></font> file system.
<p><pre> <b>per_cpu.many_util</b> High number of saturated processors</pre>
The processor utilization for at least pct percent of the CPUs
exceeded threshold percent during the last sample interval. Only applies to multi-processor systems having more than min_cpu_count
processors - for single-processor systems refer to the cpu.util rule, for multi-processor systems with less than min_cpu_count
processors refer to the per_cpu.some_util rule.
<p><pre> <b>per_cpu.some_util</b> High per CPU processor utilization</pre>
The processor utilization for at least one CPU exceeded threshold
percent during the last sample interval. Only applies to multi-processor systems with less than max_cpu_count processors -
for single-processor systems refer to the cpu.util rule, and for multi-processor systems with more than max_cpu_count processors
refer to the cpu.many_util rule. For Origin 200 and Origin 2000 systems, use the command
<br><center>$ pminfo -f hinv.map.cpu</center>to discover the abbreviated PCP names of the installed CPUs and
their corresponding full names in the <font face="Courier"><tt>/hw</tt></font> file system.
<p><pre> <b>per_cpu.syscall</b> High per CPU system call rate</pre>
The number of system calls per second for at least one CPU
exceeded threshold over the past sample interval. This rule only applies to multi-processor systems, for
single-processor systems refer to the cpu.syscall rule. For Origin 200 and Origin 2000 systems, use the command
<br><center>$ pminfo -f hinv.map.cpu</center>to discover the abbreviated PCP names of the installed CPUs and
their corresponding full names in the <font face="Courier"><tt>/hw</tt></font> file system.
<p><pre> <b>per_cpu.system</b> Some CPU busy executing in system mode</pre>
Over the last sample interval, at least one CPU was active for
busy percent or more, and the ratio of system time to busy time exceeded threshold percent. Only applies to multi-processor
systems, for single-processor systems refer to the cpu.system rule. For Origin 200 and Origin 2000 systems, use the command
<br><center>$ pminfo -f hinv.map.cpu</center>to discover the abbreviated PCP names of the installed CPUs and
their corresponding full names in the <b>/hw</b> file system.
<p><pre> <b>per_disk.util</b> High per spindle disk utilization</pre>
For at least one spindle, disk utilization exceeded threshold percent during the last sample interval.
<p><pre> <b>per_netif.collisions</b> High collision rate in packet sends</pre>
More than threshold percent of the packets being sent across an
interface are causing a collision, and packets are being sent across the interface at packet_rate packets per second.
Ethernet interfaces expect a certain number of packet collisions, but a high ratio of collisions to packet sends is indicitive of a
saturated network.
<p><pre> <b>per_netif.errors</b> High network interface error rate</pre>
For at least one network interface, the error rate exceeded threshold errors per second during the last sample interval.
<p><pre> <b>per_netif.packets</b> High network interface packet transfers</pre>
For at least one network interface, the average rate of packet
transfers (in and/or out) exceeded the threshold during the last sample interval.
This rule is disabled by default because the per_netif.util rule is more generally useful as it takes into consideration each
network interfaces' reported bandwidth. However, there are some situations in which this value is zero, in which case an absolute
threshold-based rule like this one will make more sense (for this reason it should typically be applied to some network interfaces,
but not others - use the "interfaces" variable to filter this).
<p><pre> <b>per_netif.util</b> High network interface utilization</pre>
For at least one network interface, the average transfer rate (in
and/or out) exceeded threshold percent of the peak bandwidth of the
interface during the last sample interval.
<p><pre> <b>rpc.bad_network</b> RPC network transmission failure</pre>
More than threshold percent of sent client remote procedure call
(RPC) packets are timing out before the server responds and the
number of timeouts is significantly more than the number of duplicate
packets being received (indicating lost packets).
The networked file system (NFS) utilizes the RPC protocol for its
client-server communication needs. This high failure rate when sending
RPC packets may be due to faulty network hardware or inappropriately
sized NFS packets (packets possibly too large).
<p><pre> <b>rpc.slow_response</b> RPC server response is slow</pre>
More than threshold percent of sent client remote procedure call
(RPC) packets are timing out before the server responds and the
number of timeouts is roughly equivalent to the number of duplicate
packets being received.
The network file system (NFS) utilizes the RPC protocol for its
client-server communication needs. This high timeout rate when
sending RPC packets may be because the NFS server is processing
duplicate requests from the clients which were sent after the
original requests timed out.
<p><pre> <b>espping.response</b> System Group Manager slow service response</pre>
A service being monitored by the SGI Embedded Support Partner Group
Manager has taken more than threshold milliseconds to complete, during the last sample interval. The hosts parameter specifies
hosts running the ssping PMDA, not hosts being monitored by this PMDA. The latter are encoded in the "instances" for each
espping PMDA metric - run<br><center>$ pminfo -f espping.cmd</center>to list the instances and values for the espping.cmd metric.
<p><pre> <b>espping.status</b> System Group Manager service probe failure</pre>
A service being monitored by the SGI Embedded Support Partner Group
Manager has either failed, or not responded within a timeout period (as defined by espping.control.timeout) during the last sample
interval. The hosts parameter specifies hosts running the espping PMDA, not hosts being monitored by this PMDA. The latter are
encoded in the "instances" for each espping PMDA metric - run<br><center>$ pminfo -f espping.cmd</center>to list the instances
and values for the espping.cmd metric.
<A NAME="archive_database"></A>
<hr width=100%>
<p><b>Archive Database</b></p>
<P>Use the <b>Archive Database</b> command to delete a previously archived
database or to get instructions for archiving.</P>
<UL>
<LI>The <font face="Courier"><TT>Archive</TT></font> database option conserves disk space by compressing the
current database. The SGI Embedded Support Partner can continue to read the compressed data. To ensure data
integrity, you must execute the command <font face="Courier"><TT>esparchive</TT></font> from a command line.
All Embedded Support Partner daemons are shut down during this operation and will be automatically restarted
when archiving is completed. Archiving is possible only when the size of the current database is at least
10 megabytes; the compression mechanism will not work if you try to archive a smaller database.<p>
<LI>The <font face="Courier"><TT>Delete Database</TT></font> option deletes an archived database that you no
longer need.
</UL>
<p><HR NOSHADE SIZE="3">
</BODY>
</HTML>