AutoDiscovery User`s Guide (9030727)

AutoDiscovery User`s Guide (9030727)
AutoDiscovery User’s Guide
Cabletron Systems reserves the right to make changes in specifications and other information
contained in this document without prior notice. The reader should in all cases consult Cabletron
Systems to determine whether any such changes have been made.
The hardware, firmware, or software described in this manual is subject to change without notice.
Virus Disclaimer
Cabletron has tested its software with current virus checking technologies. However, because no
anti-virus system is 100% reliable, we strongly caution you to write protect and then verify that
the Licensed Software, prior to installing it, is virus-free with an anti-virus system in which you
have confidence.
Cabletron Systems makes no representations or warranties to the effect that the Licensed
Software is virus-free.
Copyright © April 1998, by Cabletron Systems, Inc. All rights reserved.
Printed in the United States of America.
Order Number: 9030727 E12
Cabletron Systems, Inc.
P.O. Box 5005
Rochester, NH 03866-5005
SPECTRUM, SPECTRUM IMT/VNM logo, DCM, IMT and VNM are registered trademarks,
and AutoDiscovery, SpectroGRAPH, SpectroSERVER, Device Communications
Manager, Inductive Modeling Technology, Device Communications Manager, and
Virtual Network Machine are trademarks of Cabletron Systems, Inc.
UNIX is a trademark of UNIX System Laboratories, Inc.
Ethernet is a trademark of Xerox Corporation.
Solaris is a registered trademark of Sun MicroSystems, Inc.
Windows and Windows NT are trademarks of Microsoft Corporation
9030727 E12
Restricted Rights Notice
(Applicable to licenses to the United States Government only.)
1. Use, duplication, or disclosure by the Government is subject to restrictions as set forth in
subparagraph (c) (1) (ii) of the Rights in Technical Data and Computer Software clause at
DFARS 252.227-7013.
Cabletron Systems, Inc., 35 Industrial Way, Rochester, New Hampshire 03866-5005.
2. (a) This computer software is submitted with restricted rights. It may not be used,
reproduced, or disclosed by the Government except as provided in paragraph (b) of this
Notice or as otherwise expressly stated in the contract.
(b) This computer software may be:
Used or copied for use in or with the computer or computers for which it was
acquired, including use at any Government installation to which such computer or
computers may be transferred;
Used or copied for use in a backup computer if any computer for which it was
acquired is inoperative;
Reproduced for safekeeping (archives) or backup purposes;
Modified, adapted, or combined with other computer software, provided that the
modified, combined, or adapted portions of the derivative software incorporating
restricted computer software are made subject to the same restricted rights;
Disclosed to and reproduced for use by support service contractors in accordance with
subparagraphs (b) (1) through (4) of this clause, provided the Government makes
such disclosure or reproduction subject to these restricted rights; and
Used or copied for use in or transferred to a replacement computer.
Notwithstanding the foregoing, if this computer software is published copyrighted
computer software, it is licensed to the Government, without disclosure prohibitions, with
the minimum rights set forth in paragraph (b) of this clause.
(d) Any other rights or limitations regarding the use, duplication, or disclosure of this
computer software are to be expressly stated in, or incorporated in, the contract.
(e) This Notice shall be marked on any reproduction of this computer software, in whole or in
AutoDiscovery User’s Guide
Who Should Read This Guide ................................................................................................v
How to Use This Guide...........................................................................................................v
Questions about SPECTRUM Documentation ................................................................... vi
Chapter 1
The Discovery Process
What is AutoDiscovery?...................................................................................................... 1-1
The Benefits of Using AutoDiscovery ................................................................................ 1-1
Modeling and Mapping Conventions ................................................................................. 1-2
Phases of Discovery ............................................................................................................ 1-3
Phase One - The Router/Network Level ..................................................................... 1-4
Phase Two - The LAN/Bridge Level ............................................................................ 1-5
Phase Three - The Discrete LAN/Hub Level .............................................................. 1-6
Discovery Options............................................................................................................... 1-7
Duration of Discovery Sessions.......................................................................................... 1-8
Constraints ......................................................................................................................... 1-9
Chapter 2
The User Interface
Accessing AutoDiscovery.................................................................................................... 2-1
Establishing Discovery Settings ........................................................................................ 2-2
Discovery Methods ....................................................................................................... 2-3
IP Address Ranges ....................................................................................................... 2-4
Discovery Protocols ...................................................................................................... 2-5
SNMP Community Names .......................................................................................... 2-6
File Menu and Control Button Options ...................................................................... 2-7
AutoDiscovery Options ................................................................................................ 2-8
AutoDiscovery Status .................................................................................................. 2-9
Background Discovery...................................................................................................... 2-10
The Background Discovery Dialog Box..................................................................... 2-12
The Show and Hide Dialog Box................................................................................. 2-14
Chapter 3
Using AutoDiscovery
Before You Start.................................................................................................................. 3-1
9030727 E12
Creating the Initial Topological Model ..............................................................................3-1
Additional Discovery and Configuration Tips ...................................................................3-4
Clustering Networks ....................................................................................................3-4
Appendix A
IP Addresses
IP Address Structure ......................................................................................................... A-1
Class A Networks ........................................................................................................ A-1
Class B Networks ........................................................................................................ A-2
Class C Networks ........................................................................................................ A-2
Deciphering an IP Address.......................................................................................... A-2
Subnets and Subnet Addresses ......................................................................................... A-3
Appendix B
AutoDiscovery from the Command Line
Entering Startup Commands ............................................................................................ B-1
Using crontab Scripts ........................................................................................................ B-2
SPECTRUM Schedule Manager ................................................................................. B-4
Appendix C
AutoDiscovery User’s Guide
This document provides instructions for using AutoDiscovery, a SPECTRUM core application
program that operates in conjunction with the SPECTRUM network management software to
create a model of an existing network.
Who Should Read This Guide
This guide is intended for SPECTRUM administrators and technicians
responsible for determining network configuration and overseeing network
operations. By following the instructions and procedures described herein, the
reader will be able to use AutoDiscovery to create and maintain an accurate
network model that will facilitate full exploitation of SPECTRUM’s powerful
management and monitoring capabilities.
This guide assumes that the administrative user is experienced with
SPECTRUM and its administration as described in the SPECTRUM
Administration documentation. Any user should also be familiar with the
SpectroGRAPH™ user interface, and with the user functions explained in the
SPECTRUM Operation documentation.
How to Use This Guide
This guide contains both general information and detailed instructions. The
document is organized as follows:
• Chapter 1 provides an overview of AutoDiscovery and describes the
process it uses to construct a network model. This chapter also identifies
the limitations and constraints within which the application functions.
• Chapter 2 explains how to use the AutoDiscovery dialog box to establish
network bounds and other settings for a discovery session.
9030727 E12
Questions about SPECTRUM Documentation
• Chapter 3 provides step-by-step instructions for running AutoDiscovery
and suggests strategies for configuring your network topology views after
discovery operations are completed.
• Appendix A discusses IP address conventions and subnetting schemes.
• Appendix B explains how to run AutoDiscovery via UNIX crontab scripts.
This lets you take advantage of automatic discovery capabilities on a
continuous or periodic basis.
• Appendix C is a glossary of the technical terms and acronyms used in this
Questions about SPECTRUM Documentation
Send your questions, comments or suggestions regarding SPECTRUM
documentation to the Technical Communications Department directly via the
following internet address:
[email protected]
AutoDiscovery User’s Guide
Chapter 1
The Discovery Process
This chapter provides an overview of AutoDiscovery and describes the methodology it uses to
construct a network model. Constraints and limitations that apply to the process are identified
along with certain variables that can affect the duration of a discovery session.
What is AutoDiscovery?
AutoDiscovery is a key part of SPECTRUM. It is an application program that
virtually automates the creation and maintenance of the software model
SPECTRUM uses to manage your network. Operating within IP address
ranges and other guidelines that you specify, AutoDiscovery explores your
network and creates individual models of the devices and other network
entities it finds. These models are stored in the VNM database along with
information about their relationships and interconnections. SPECTRUM’s
graphical user interface, SpectroGRAPH, provides access to the overall
network model through a hierarchy of Topology views in which AutoDiscovery
places icons representing each of the network elements that are modeled in
the database.
The Benefits of Using AutoDiscovery
By automating most of the modeling process, AutoDiscovery can save
significant amounts of time for SPECTRUM users, especially in the case of
very large networks. However, beyond the time saved in finding devices,
creating models, and placing icons in the topological representation of your
network, AutoDiscovery can enhance the effectiveness of SPECTRUM itself.
For one reason, AutoDiscovery performs a very comprehensive exploration; it
will discover elements you might well miss if attempting to manually model
your network. Furthermore, the network model AutoDiscovery creates is
specifically designed to accommodate the particular way that SPECTRUM
works. AutoDiscovery’s placement of models within the topological scheme
9030727 E12
Modeling and Mapping Conventions
supports SPECTRUM intelligence and optimizes its ability to identify
problem areas and accurately reflect current conditions.
The end result of the AutoDiscovery process is a stratified abstraction of your
network that lets you access information at the precise level of detail you
need. This makes it easier for you to make critical management decisions
whether on a high-level, network-to-network basis, or down at the hub port or
endpoint device level.
Modeling and Mapping Conventions
As AutoDiscovery “discovers” devices and other network entities, it creates
the appropriate SPECTRUM models and adds them to the SpectroSERVER
database. AutoDiscovery also creates icons to represent these entities (and the
logical connections* between them) and by default places the icons in the
appropriate SPECTRUM Topology views according to the rules described
below. Some of these defaults can be overridden through the Discovery
Options described on Page 1-7.
1. Icons representing routers that connect two or more IP Class A, B or C
networks are placed at the Universe (top) level of the SPECTRUM
topology scheme.
2. Icons representing IP Class A, B and C network models are placed at the
Universe level.
3. If a router connects subnets within an IP Class A, B, or C network, the
router icon is placed within the Topology view for that network’s model.
Each subnet connected by such a router is modeled as a LAN, and the
corresponding icon is placed within the same view. Note that a single LAN
model may embrace multiple subnet ranges. AutoDiscovery automatically
combines multiple addresses into a single LAN when routers are
configured with multiple addresses on an interface. Note also that if a
router within an IP Class A, B, or C network has only a single non-serial
interface, it will be modeled within the LAN associated with that
interface. WA_Link models are created for wide-area type interfaces (T1,
T3, X.25, etc.) If a wide-area interface has multiple subnets associated
with it, a separate WA_Link model will be created to represent each
logical connection.
4. Icons for bridges and for the discrete LANs they interconnect are placed
within the Topology view for the model of the LAN that they comprise.
(Discrete LAN model types include: 802.3, 802.5, FDDI, etc.)
5. Icons for hub and fanout models are placed within the Topology view of
the discrete LAN model to which they belong. The fanout model type is
used to model any non-intelligent or unidentifiable device that provides
connectivity within a LAN. Examples of fanouts include coaxial cable
segments, media access units (MAUs), and multiport transceivers.
The Discovery Process
AutoDiscovery User’s Guide
Phases of Discovery
6. Endpoint device icons are connected to the appropriate port within a
Device Topology view, if the device’s address is the only one heard on that
port. Icons for endpoint devices connected to fanouts, are placed within
the fanout model’s Cablewalk view or Cablewalk List view.
* By default, AutoDiscovery creates gold or silver “pipe” icons to indicate
logical connections between network entities. A connection is said to be
“resolved” when SPECTRUM can determine specific device ports at both ends.
The pipe icon is colored gold when a connection is resolved, silver when the
connection has not been resolved. For example, if AutoDiscovery places a gold
connection pipe between a router model icon and a LAN model icon, one of the
ports in the router’s DevTop view will show a connection to the LAN and to
some other device within that LAN. Likewise, the other device’s DevTop view
will show a connection to the router (and to the network group entity that
contains the router). You can also draw connections between icons manually,
and SPECTRUM will attempt to resolve them. However, unresolved userdrawn connections may be erased by subsequent AutoDiscovery sessions if the
connections do not comply with the modeling and mapping conventions
followed by SPECTRUM.
SPECTRUM 5.0 also features a resource-conserving Live Pipes service that
you can enable/disable for the VNM (server) model as a whole, then toggle on
and off as desired for selected connections. When this service is enabled, a
different range of pipe colors is used to indicate the current status of the link
(good, bad, disabled, unknown, or unreachable). For further information and a
key to the Live Pipes status color code, refer to Getting Started for
Administrators or How to Manage Your Network with SPECTRUM.
Phases of Discovery
As the modeling and mapping conventions discussed in the previous section
suggest, AutoDiscovery is designed to operate in a hierarchical manner. That
is, it explores the network and populates SPECTRUM Topology views from
the top down. The largest or most general network groups are placed within
the Universe level Topology view with subnets and devices appearing in
successively lower level views until, finally, endpoint devices are shown
connected to specific ports in DevTop views or to cable segments in Cablewalk
The scope of this process depends upon the IP address ranges and other
guidelines you establish using the main AutoDiscovery dialog box described in
Chapter 2. It also depends on the particular level of the topological hierarchy
from which you are running the AutoDiscovery application. When you start at
the Universe level using all of the available discovery methods and protocols
(and a sufficiently broad address range), discovery proceeds automatically,
and the entire network can be explored and mapped in a single session. In
practice, however, and especially for larger networks, it is often preferable to
first run the application at the Universe level with only the Router Discovery
9030727 E12
The Discovery Process
Phases of Discovery
Phase One - The Router/Network Level
method selected, so that discovery proceeds only to the level governed by
Rules 1 through 3 (i.e., mapping only routers, IP Class A, B, or C networks,
LANs and wide area links). AutoDiscovery can then be run separately for each
of the discovered LANs. This approach allows the administrator to more
closely monitor the process and facilitates isolation and resolution of any
anomalies or problems encountered. In any case, the complete discovery
process occurs in three basic phases, which are described in the following
sections as if they are occurring in a single session.
Phase One - The Router/Network Level
AutoDiscovery always begins by checking the SPECTRUM database for
existing models of “seed” routers whose route tables can be used as a source of
information to “grow” the initial high-level topology. For all such routers that
are currently responding (and whose models are not currently in
SPECTRUM’s Lost and Found View), AutoDiscovery reads the route table to
identify other routers listed as next hops. The next hop addresses are
examined in terms of the search range established for the discovery session.
Those that are within the range (or that are associated with destination
addresses within the range) are added to the list of routers that will be
processed in this same manner. If no model currently exists for the next hop
address, AutoDiscovery creates one. Subnets that the table indicates are
associated with a particular interface are stored in the LAN model associated
with that interface. Again, AutoDiscovery creates the LAN model if it doesn’t
already exist.
As routers on the list are being processed, AutoDiscovery creates the
appropriate network models (IP Class A, B, or C) and populates them with
models of the discovered LANs and wide area links, placing router models at
the same topological level as the entities they connect. Placement of models is
accomplished using the interface and network mask and class information
obtained from the route tables and is done in accordance with the modeling
and mapping conventions previously described.
Figure 1-1 is an example of the topology mapping that might result from
Phase One discovery. The Universe-level Topology View in the top part of the
figure shows two IP Class B networks connected by a router. This router is
shown at the top level because it is a border router (i.e., one that connects two
or more distinct IP networks). By opening the Topology View for either of the
Class B network icons (New York or Chicago), you can view the contents of
these networks as shown in the lower portion of the figure. The view on the
left contains icons representing three LANs connected via an interior router
called Router 2. This router is connected by a wide area link back to Router 1,
which is represented by a triangular off-page reference icon. Likewise, the
view on the right shows Router 1 as an off-page reference icon. Since Router 1
actually connects the two Class B networks, it is more accurate to show its
device icon at the Universe level between the network icons rather than inside
the views for those networks.
The Discovery Process
AutoDiscovery User’s Guide
Phases of Discovery
Phase Two - The LAN/Bridge Level
Figure 1-1.
Typical Phase One Discovery Results
Primary Landscape 0x40000 VNM topanga Universe of type Universe
File View
New York
IP Class B
Chicago of IP Class B
New York of IP Class B
File View
LAN #1
IP Class B
Router 1
File View
Router 2
LAN #6
LAN #2
LAN #4
LAN #5
LAN #3
See Figure 1-2 for the view from LAN #3
Phase Two - The LAN/Bridge Level
The second phase of discovery occurs at the LAN/bridge level. Here
AutoDiscovery uses one or more of three user-selectable discovery methods* to
examine each of the LANs discovered in Phase One. In the course of
examining each LAN, AutoDiscovery locates and models all supported bridges
and uses their interface information to model and place discrete LANs (802.3,
802.5, etc.) that the bridges interconnect. If no bridges are found, hubs will be
examined, and if any hub shows an interface type corresponding to a discrete
9030727 E12
The Discovery Process
Phases of Discovery
Phase Three - The Discrete LAN/Hub Level
LAN model type, then a discrete LAN model will be created. The discrete LAN
model will collect all hubs showing the same interface type if they are not
already collected by another discrete LAN model.
The top part of Figure 1-2 shows the results of Phase Two discovery for LAN
#3, which was part of the “New York” IP Class B network-level view shown in
Figure 1-1. This view shows that LAN #3 contains three 802.3 discrete
networks, one 802.5 discrete network, and two bridges. There is also an offpage reference icon to show the connection between LAN 802.3 #2 and Router
2, whose device icon was placed in the New York network view in Figure 1-1.
*The Range-Test, NIS, and Address Resolution Table discovery methods are
selectable from the main AutoDiscovery dialog box, which is discussed in
detail in Chapter 2 of this guide.
Phase Three - The Discrete LAN/Hub Level
In the third phase of discovery, the selected discovery methods are used to
examine each of the discrete LANs discovered in Phase Two. As each hub is
located and modeled, AutoDiscovery also attempts to identify and model the
devices connected to each of its ports. (As noted in Item 6 under “Modeling and
Mapping Conventions,” a model will be placed on a hub or bridge port only if
the model’s address is the only one heard on the port.) Other non-intelligent
(or unidentifiable) multiport devices, such as a multiport transceiver with
several users attached, are modeled as fanouts.
The lower view in Figure 1-2 shows the results of Phase Three discovery for
the discrete LAN 802.3 #2. Here three hubs and two fanouts have been
discovered. Two off-page reference icons also appear in the view, one
representing the connection to Bridge 1 at the next higher topological level
and the other showing the connection to Router 1, whose device icon appears
in Figure 1-1.
As the figures illustrate, when all three phases of discovery have been
completed, the result is a comprehensive, topologically accurate
representation of your network that lets you access the level of detail you
want, from major network groups down to endpoint devices.
The Discovery Process
AutoDiscovery User’s Guide
Discovery Options
Figure 1-2.
Typical Discovery Results for Phases Two and Three
LAN #3 of type LAN
File View
802.3 #1
(created by
Phase Two)
Bridge 1
802.3 #3
Bridge 2
802.3 #2
802.5 #1
#2 of type LAN 802.3
File View
(created by
Phase Three)
Bridge 1
Fanout #1
HUB #1
Fanout #2
HUB #2
HUB #3
Discovery Options
Some of AutoDiscovery’s modeling and mapping conventions (see Page 1-2)
can be selectively overridden for Universe, Network, and LAN models through
the Discovery Options panel that appears in the associated Information
view. If you are already in the Topology view of the model for which you want
to access these options, select View > Current View Information. If you are
9030727 E12
The Discovery Process
Duration of Discovery Sessions
in the Topology view that contains the model, highlight the model’s icon, then
select View > Icon Subviews > Model Information. The location of the
Discovery Options panel within the Information view varies depending on
which model type is selected, but in each case the panel provides the following
three menu buttons, each of which lets you select an alternative to the default
way in which models will be created and placed during AutoDiscovery.
AutoPlace Routers
The default setting for this button will automatically place routers according
to the modeling and mapping conventions described on Page 1-2.
Alternatively, you can toggle this button to the Place Routers in this view
only option, which will place any newly discovered routers in the view from
which you invoked AutoDiscovery.
Create IPClass models
The default setting for this button will create models for IP Class A, B and C
networks and place the icons in the Universe Topology view. You can suppress
creation of IP Class models by toggling this button to the Place LANs in this
view only option, which will create LAN models instead of IP Class models
and place them only in the view from which you invoked AutoDiscovery.
Create LAN models
The default setting for this button will create and place LAN models according
to the conventions described on Page 1-2. You can suppress creation of LAN
models by toggling this button to the Don’t Create LAN models option,
which will prevent creation of new LAN models during Phase One discovery.
Duration of Discovery Sessions
The amount of time required to run AutoDiscovery depends on many
variables, including:
The magnitude of the network address range(s) you specify
The number of discovery methods you choose
The number of discovery protocols you choose
The number of community names you specify
The subnetting scheme in use
The number of protocols AutoDiscovery must use to identify a device
The number and type of nodes in your network
Given the number of variables involved, it is impossible to accurately predict
discovery time based solely on the number of nodes in a network. The time
The Discovery Process
AutoDiscovery User’s Guide
needed for a complete network discovery can vary from minutes to hours,
depending on the nature of the network involved.
The network map created by AutoDiscovery can only be as accurate as the
information available from the network devices. Since this information is
typically derived from memory caches within a device, some information may
be missing from a device at the time AutoDiscovery contacts it. In other cases,
the information available from two different devices may be in conflict,
especially right after a change in configuration. Thus you may not discover
every device on your network the first time you run AutoDiscovery, and you
should consider running additional discovery sessions to further refine your
network model with the most recent device data. Indeed, if the configuration
of your network changes frequently, AutoDiscovery should be run at regular
intervals, either using the Background Discovery feature described in Chapter
2, or as a UNIX cron job as described in Appendix B. Each succeeding session
will improve the accuracy and completeness of your model.
Beyond the issue of device status at any given moment, you should keep in
mind that AutoDiscovery can only map correctly using repeaters, bridges, and
routers for which management protocols provide the necessary port
connectivity information (addresses heard on each port). Moreover, the
appropriate SPECTRUM management modules must be available to provide
access to that information.
You should also be aware that AutoDiscovery cannot remove devices from the
database nor from the visible topological model, since there is no positive
indication that a device has been removed from the network. A device that can
not be contacted may be powered off, temporarily out of service, or simply not
transmitting. Therefore, if a device is permanently removed (physically) from
your network, you must manually destroy the device’s model from the
database. Otherwise, the model will continue to generate a Lost Contact
alarm (condition color red).
The following device-specific constraints and limitations also apply:
• Each router interface to a given network must have the same subnet mask
(see Appendix A for more information on subnet addresses).
• Completely redundant routers may not be detected unless either the
Range-Test or NIS discovery methods are used (see Chapter 2 for more
information on discovery methods).
• If a combination of routing protocols imposes artificial fragmentation on a
network, each such fragment should have its own “seed” router to
facilitate full discovery of its contents.
• Redundant Cabletron SNMP bridges are supported and will be placed in
the proper LANs, but the views will require editing to form adjacencies.
9030727 E12
The Discovery Process
This is because bridges in standby mode have unreliable bridging tables,
and AutoDiscovery will not be able to determine the proper port
• Bridges with long timeout periods for their bridging tables may continue
to generate information for devices that have since been removed.
• Although IRBMs are not supported as bridges, AutoDiscovery will
properly find and place these devices as hubs.
• In spanning tree protocol, a bridge that is a leaf on a spanning tree only
broadcasts spanning tree information on the port connected to the
“upstream” bridge. Thus a “downstream” hub may never hear any
transmissions using the MAC address of an adjacent bridge, and the
topology generated by AutoDiscovery may not show the hub-bridge
• Discovery of hubs may generate physical address models in the Lost and
Found View.
The Discovery Process
AutoDiscovery User’s Guide
Chapter 2
The User Interface
This chapter explains how to access and operate the dialog boxes that the user interface for the
AutoDiscovery application.
Accessing AutoDiscovery
AutoDiscovery can be run either from within SpectroGRAPH or from your
operating system’s command line (see Appendix B for command format).
Either method provides access to the user interface through which you
determine the boundaries of the network to be discovered and the type of
discovery and mapping that will be performed. Once settings are established,
you can run subsequent discovery sessions automatically at predetermined
intervals using either the Background Discovery feature described on
Page 2-10, or a UNIX “cron” facility as explained in Appendix B.
To invoke AutoDiscovery from SpectroGRAPH, select the Auto Discover
option from the Edit menu of any SPECTRUM Topology view for which
discovery would be applicable (i.e., above the level of discrete LAN). This will
display the main AutoDiscovery dialog box shown in Figure 2-1. Note that the
title bar identifies the applicable landscape handle and VNM name. There is
also a subtitle indicating the Model Name and Model Type of the model from
which you invoked AutoDiscovery.
9030727 E12
Establishing Discovery Settings
Figure 2-1.
The AutoDiscovery Dialog Box
AutoDiscovery - Landscape 0x40000 VNM topanga
Model Universe of Type Universe
Discovery Methods
Discovery Protocols
Router Discovery
LAN Discovery
Range Test Discovery
NIS Discovery
ARP Table Discovery
SNMP Community Names
IP Address Ranges
Community Name:
Establishing Discovery Settings
The dialog box shown in Figure 2-1 lets you control and focus the discovery
process by determining the discovery method(s) to be used and the protocols
and address ranges to be considered. These settings can be saved (via the File
menu) for default use in subsequent discovery sessions. Individual selection
buttons and data entry fields are described on the following pages.
The User Interface
AutoDiscovery User’s Guide
Establishing Discovery Settings
Discovery Methods
Discovery Methods
Whenever you run AutoDiscovery, it performs Phase One discovery to identify
routers, LANs and wide area links (see Chapter 1 for an overview of the three
phases of discovery), thus the Router Discovery button is permanently set to
“on” by default. Subsequent actions, however, depend upon the other discovery
methods that you select. The first time you run AutoDiscovery, you
should do so at the Universe level with Router Discovery the only
method selected. For subsequent sessions, however, you can activate any of
the following three discovery methods by clicking on the associated selector
button. Note that if none of these discovery methods is selected,
AutoDiscovery will stop after Phase One and will not attempt to discover
details for each LAN.
Router Discovery
When this method is selected, AutoDiscovery searches the SPECTRUM
database for a “seed” router model that it can use as the basis for Phase One
discovery (see Chapter 1 for an overview of the three phases of discovery). As a
default, Router Discovery uses the seed router’s IP Routing Table to find and
model other routers listed as next hops. Alternatively, you can specify that the
router’s IP Address Table be used (see AutoDiscovery Options on Page 2-8). In
addition to router models, Router Discovery creates the appropriate IP Class
A, B, or C models and populates them with LAN and WAN models according
to the modeling and mapping conventions described in Chapter 1.
LAN Discovery
This method enables the mapping of existing device models (e.g., those
already discovered during a Background Discovery session) at the LAN/bridge
level and below. Although you can run AutoDiscovery with only this method
selected, it is automatically selected whenever any of the following three
methods (Range Test, NIS, or ARP Table) are selected.
Range Test Discovery
When this method is enabled, AutoDiscovery uses ICMP echo requests (pings)
to test each of the IP addresses within the range or ranges you specify in the
IP Address Ranges panel explained below. An address that responds to a ping
is then tested against selected protocols until it can be identified and the
appropriate model created. Although this method provides comprehensive
coverage of a given range, the tradeoff in terms of bandwith usage should be
considered before using it on larger ranges.
NIS Discovery
(Solaris systems only)
Unless another method is also selected, this method limits discovery to those
devices addressed in the host table for your Solaris system’s NIS (Network
Information Service) server. Although this method is not dependent on any
9030727 E12
The User Interface
Establishing Discovery Settings
IP Address Ranges
particular protocol, subsequent identification of devices is facilitated by
having both the available Discovery Protocols selected.
ARP Table Discovery
This method allows AutoDiscovery to associate a discovered device’s IP
address to a physical (MAC) address so that resolution to the device port level
can be achieved duringPhase Three discovery. To do this, AutoDiscovery
references the Address Resolution Protocol (ARP) Table of each modeled
router and tests any IP addresses within the specified range. For models that
already exist in the SpectroSERVER database, the MAC_Address attribute is
updated with the physical address associated with the corresponding IP
address in the table. If no model exists for a table entry, AutoDiscovery will
attempt to create the appropriate model based on the selected protocol(s). For
example, if the device does not respond to SnmpPif but does respond to
IcmpPif, a pingable model will be created.
The ARP Table method of discovery should not be used as the sole method,
since models would not be created for any devices that do not appear in the
ARP tables of your routers (such as bridges and hubs). Note also that this
method will not produce any results if there are no routers within the
specified address range(s).
IP Address Ranges
This panel allows you to set the bounds of your network by specifying one or
more IP address ranges within which discovery operations will be confined.
The top section of the panel (adjacent to the scroll bar) lists the ranges that
will be tested when you actually start the application. Each line in the list
defines a range by a low address and a high address. For instance, in the
example shown by Figure 2-1, AutoDiscovery will test addresses between and
In the lower portion of this panel there are From and To data entry fields. To
add a new range to the list, click within the From field on the left and type in
the IP address for the low end of the desired range. Then tab to or click within
the To field on the right and enter the address for the high end. Finally, click
on the Add button at the bottom of the panel to transfer your entry to the list.
To delete a range from the list, click on the desired range to highlight it, then
click on the Delete button. (Note that the Delete button is disabled unless
one or more ranges are selected.) To modify a range in the list, double-click on
it to move it down to the data entry fields, then edit as needed and add it back
to the list.
The first time you run AutoDiscovery at the Universe level, no default ranges
will have been established yet, so you will have to enter at least one range. A
single range is appropriate if your network’s address bounds are contiguous;
however, if your network has multiple IP Class A, B, and C networks, you will
The User Interface
AutoDiscovery User’s Guide
Establishing Discovery Settings
Discovery Protocols
need to specify multiple address ranges. Also, if you are sure that your host
addresses are limited to a subset of a full network range, discovery will
proceed faster if you specify only those ranges currently in use.
You can save the bounds and other settings in this dialog box by pulling down
the File menu and selecting the Save Current option. This will establish the
settings as defaults for subsequent executions of AutoDiscovery and is
especially useful if you plan to use the Background Discovery feature (see
Page 2-10), or to run AutoDiscovery at regular intervals via a UNIX cron job
(see Appendix B). Note that address range bounds need only be set at the
Universe level. If you later invoke AutoDiscovery for a network entity
discovered during an initial discovery session (IP network, LAN, wide area
link), this panel will show the appropriate bounds. To speed the discovery
process, however, you may want to narrow some of these bounds to encompass
only the portions of these ranges currently in use.
As noted above, when running AutoDiscovery at lower topological levels, you
can specify subsets of the address range(s) used for Universe-level discovery,
but you cannot use addresses outside of the range or ranges that have been
saved either at the Universe level or for the lower level. In other words, you
can scope down but not up. For example, assume the IP address range saved
at the Universe level is through If you then
try to run AutoDiscovery for a Class B network and enter a range
of through, you will get an “address out of
bounds” error message since the 124 subnet is not within the saved range.
When AutoDiscovery is run from the command line, user-entered ranges for
a given LAN will by default override the ranges established for that LAN
when AutoDiscovery was run at the Universe level. To prevent the originally
established ranges from being further restricted, you can add the argument
-nolanrestrict to the startup command. See Appendix B for further
information on command format.
If you are running AutoDiscovery on a network or a part of a network with
no routers, be sure that the IP address range(s) you use contain no broadcast
addresses, since all devices will respond to the ping, and a device model may
be created for whichever one responds first. The resultant model will then
regularly poll the broadcast address, which may degrade network
performance and have unpredictable results on SPECTRUM operations.
Discovery Protocols
Select one or both of the discovery protocols listed in this panel by clicking on
the protocol name(s). When a protocol is selected it appears highlighted. You
can also deselect a highlighted protocol by clicking on it.
9030727 E12
The User Interface
Establishing Discovery Settings
SNMP Community Names
AutoDiscovery uses the selected protocols in its attempts to contact and
identify a network entity at each IP address in the specified range(s).
Protocols are tried in the same order in which they are listed in this panel,
and AutoDiscovery categorizes the entity according to the first protocol to
which it responds. Thus, a device capable of responding to both SNMP and
ICMP protocols will be identified by the SNMP protocol if both are selected.
Each protocol used adds to overall discovery time, so you should try to select
only those protocols that you actually need for your network. However, if you
do not select any protocols, AutoDiscovery will yield a large number of
physical address and pingable models, rather than models of the actual
devices. For example, if you are using the NIS discovery method, you should
select all the available protocols to obtain the best results. The protocols
currently available are:
Select this protocol to locate SNMP-compliant devices.
When this protocol is selected, the SNMP Community
Names box is activated. You must specify at least one
SNMP community name in this box. The default name is
Select this protocol when using the Range-Test method to
discover devices that do not support SNMP management
but that do support the ICMP echo function.
SNMP Community Names
Community names are passwords assigned to individual SNMP devices to
control access. If you have selected SnmpPif as a discovery protocol, this panel
lets you specify one or more SNMP community names that will be searched for
during the discovery process. Use the default name “public” if you have not
configured your SNMP devices with community names. To add a different
name, click within the Community Names data entry field, type the name,
then click on the Add button. To delete a name from the list, click on the name
to select it, then click on the Delete button. To modify a name in the list,
double-click on it to move it down to the data entry field, then edit as needed
and add it back to the list. Note that if you delete the community name
“public” but have selected the IcmpPif protocol in addition to SnmpPif, then
“public” devices will still be discovered and modeled as pingables.
The discovery process searches your network for community names in the
order in which they appear in this panel. The first community name a device
responds to is the one that is used for the model. If your devices support
multiple community names, and you have a preference as to naming, be sure
to list names in order of preference.
The User Interface
AutoDiscovery User’s Guide
Establishing Discovery Settings
File Menu and Control Button Options
File Menu and Control Button Options
With the exception of the Settings option, which is accessed only from the
AutoDiscovery dialog box’s File menu, you can select the following options
either from the menu or by using the buttons at the bottom of the dialog box.
This menu option/button starts AutoDiscovery with the
currently displayed settings and brings up the
AutoDiscovery Status window shown in Figure 2-3. Note
that once AutoDiscovery starts, the main dialog box is
disabled except for the Stop and Status buttons described
This menu option/button lets you stop an AutoDiscovery
session that is already in progress. Any discovery results
to that point will be saved.
This File menu option accesses a submenu with the
following options.
Save Current. Saves all currently specified settings from
the AutoDiscovery and AutoDiscovery Options dialog
boxes to the VNM database so that you can recall them for
future discovery sessions. With the exception of IP address
ranges set at the Universe level, these settings apply only
to the network model represented by the current view, i.e.,
the view from which you invoked AutoDiscovery. Settings
may differ for different network or LAN models.
Restore Originals. Overwrites the current settings with
the most recently saved settings recalled from the VNM
Restore Defaults. Replaces current settings with default
settings. There are two sets of default settings: those that
apply at the Universe level and those that apply at other
Universe-level defaults include the following: no discovery
methods selected, IP Address ranges blank, all protocols
selected, SNMP community name of “public.”
Defaults for lower topology levels are the same as the
settings you specified at the Universe level, with the
exception of IP Address Ranges. Default ranges at levels
other than the Universe level are taken from the IP
address list maintained by the model that discovery is run
against. As noted previously, however, the IP ranges you
specify at the Universe level will bound discovery at all
other levels. If a subnet is out of range of Universe-level
bounds, it cannot be discovered.
9030727 E12
The User Interface
Establishing Discovery Settings
AutoDiscovery Options
Figure 2-2.
This menu selection/button displays the AutoDiscovery
Options window shown in Figure 2-2.
This menu selection/button also brings up the
AutoDiscovery Status window shown in Figure 2-3.
Click on the Exit button to close the dialog box without
running AutoDiscovery. Both the Options dialog box and
the Status window provide a “Close” function as the sole
option under their File menus. This option closes the
dialog box but does not exit AutoDiscovery.
The AutoDiscovery Options Dialog Box
AutoDiscovery - Landscape 0x400000 - VNM topanga
AutoDiscovery Options
Save Database before discovering
Mapping Options
AutoPlace discovered Networks
Router Discovery Type
Ping timeout:
IP Routing Table
Ping Retry count:
AutoDiscovery Options
This dialog box allows you to further configure and control the AutoDiscovery
process through the menu buttons and data entry fields described in the
following paragraphs. To dismiss the dialog box, pull down its File menu and
select the Close option.
Save Database before discovering. Default setting is False. Set the
selector to True to ensure that a backup copy of your current database is
made before discovery operations begin.
Mapping Options. The Hierarchical option, which is the default, uses a
standard IP hierarchical mapping that places models for IP Class A, B, and C
networks and the routers connecting them in the Universe-level Topology
view, LANs and their connecting routers in a network-level view, discrete
LANs in a LAN-level view, and so on. For networks that contain many bridges
The User Interface
AutoDiscovery User’s Guide
Establishing Discovery Settings
AutoDiscovery Status
and few or no routers, the Flat option will place all discovered elements into a
fanout model within the current view. The elements can then be viewed
through the fanout model’s Cablewalk view or Cablewalk List view. This
option is mainly intended for use within LANs modeled during Phase One
AutoPlace discovered Networks. Use the None setting if you wish to
arrange discovered elements manually by dragging them with the mouse once
the discovery session is over. In this case, AutoDiscovery will simply cascade
the icons in a single overlapping group within the Topology view. The other
two settings will automatically arrange the icons in either a Radial or Tree
arrangement once the Topology view is taken out of Edit mode. Radial is the
default setting.
Router Discovery Type. When the Router Discovery method has been
selected, this menu button lets you determine whether the discovery process
will be based on reading IP Routing Tables or IP Address Tables. Use the
default option, IP Routing Table, if your primary objective is complete and
comprehensive mapping; use the IP Address Table option if you are more
concerned with reducing the amount of time required.
Ping Timeout. This field specifies the number of milliseconds per try that
AutoDiscovery will wait for an IP address to respond to a ping. The default
value is 1000.
Ping Retry Count. The value in this field specifies how many times
AutoDiscovery will attempt to ping each IP address. The default is 2.
AutoDiscovery Status
The AutoDiscovery Status window lets you monitor the progress of the
discovery session in two ways. The arrow symbols in the top panel light up
(i.e., the arrow appears white or as a light color on a darker background) to
indicate which of the adjacent stages of discovery is currently in progress,
while the scrollable AutoDiscovery Progress Log in the lower part of the
window provides detailed information on the number and type of devices
being discovered. Note that the top panel also has four fields that display the
model handle of the entity or device being processed.
9030727 E12
The User Interface
Background Discovery
Figure 2-3.
The AutoDiscovery Status Window
AutoDiscovery - Landscape 0x400000 - VNM topanga
AutoDiscovery Status
Discovering Routers and Subnetworks
Exploring Subnetworks
Testing Subnetwork Address Ranges
Scanning NIS Database
Scanning Router Address Resolution Tables
Waiting for Modules to Activate
Examining Bridge at IP Address
Mapping LANs
Reading Concentrator Source - Address Tables
AutoDiscovery Progress Log
The information displayed in the Progress Log panel is also stored in the
AutoDisc.logs subdirectory of the SG-Tools directory. Individual log files are
labeled ADISC.OUT and use extensions that identify the machine name,
user name, date and time. For example:
Background Discovery
After running AutoDiscovery on a given range of IP addresses, you can
continue the discovery process through the Background Discovery option,
which is available from the Topology view’s Edit menu for Universe, Network,
and LAN models. Background Discovery runs at intervals of your choosing
and will attempt to discover devices at IP addresses that could not be
contacted during the AutoDiscovery session. If you have not yet run
AutoDiscovery against a particular Network or LAN model, you can still use
Background Discovery as long as a Network Address and Subnet Mask were
The User Interface
AutoDiscovery User’s Guide
Background Discovery
were specified and saved for that model when it was created (whether
manually or via AutoDiscovery). If so, Background Discovery will generate a
list of addresses that are within the range for that model, and will then
attempt to contact each address. In either case, you can control the scope and
frequency of Background Discovery through the dialog boxes shown in
Figure 2-4 and Figure 2-5.
To configure and run Background Discovery against a particular Universe,
Network, or LAN model, perform the following steps:
1. With the Topology view of the Universe, Network, or LAN model to be
discovered in Edit mode, select Edit > Background Discovery. This will
display the Background Discovery dialog box shown in Figure 2-4.
2. Enter a numeric value in the Time Interval field and use the adjacent
menu button to select the desired unit (days, hours, or minutes).
3. If you want models to be created for IP addresses that could only be
contacted via ICMP pings, set the Pingables? menu button to Yes.
4. Click the IP Address button to open the Show and Hide dialog box
(Figure 2-5), which contains two scrollable lists of IP addresses. The
Undiscovered Address list shows all the addresses that Background
Discovery will attempt to contact and model. The Excluded Address list
shows all other addresses within the applicable range.
5. Use the Filter/Search feature and arrow buttons to scan the lists and
move addresses between them as desired.
6. Click Cancel to dismiss any changes you have made or click OK to
confirm the current list contents and go back to the Background Discovery
dialog box.
7. Click in within the Community Names panel, then edit to add or remove
entries as needed.
8. When you are satisfied with all settings within the dialog box, toggle the
Background Discovery menu button to select the On option. (This will
enable Background Discovery without actually starting the first session.)
9. From the File menu, select Save All Changes to start the Background
Discovery session. The current date and time will be displayed in the
Background Discovery dialog box’s Start field, and the Status field will
display the message “Running.”
Once started, Background Discovery will attempt to contact each of the IP
addresses in the Undiscovered Address list. Whenever a model is created
for a particular address, the address is automatically moved to the Excluded
Address list. The discovery session ends when all addresses on the
Undiscovered list have been tried. If any addresses remain on this list at the
end of the session, Background Discovery will automatically restart on behalf
of this model at the time interval you have specified, unless you manually
disable Background Discovery by toggling the On/Off button mentioned in
Step 8 above. The functionality of both the Background Discovery and Show
9030727 E12
The User Interface
Background Discovery
The Background Discovery Dialog Box
and Hide dialog boxes is discussed in more detail immediately following the
corresponding figures.
Figure 2-4.
The Background Discovery Dialog Box
Universe of type Universe - Landscape 0x400000 - VNM
* File
Help ?
Background Discovery
Background Discovery
Time Interval
IP Address List
Network Address
Network Mask
Started Thu May 14 13:42:03 1998
Ended Thu May 14 13:42:08 1998
Community Names
The Background Discovery Dialog Box
Accessible from the Edit menu of any Universe, Network, or LAN Topology
view (select the Background Discovery option), this dialog box lets you
control the scope and frequency of Background Discovery sessions for the
associated model. Dialog box components are discussed individually below.
Background Discovery
Off / On
Toggle this menu button On or Off to enable/disable Background Discovery
for the selected model. Selecting Save All Changes from the File menu will
The User Interface
AutoDiscovery User’s Guide
Background Discovery
The Background Discovery Dialog Box
start the discovery session only if this button is set to On. When no addresses
remain on the Show and Hide dialog box’s Undiscovered Address list
(Figure 2-5), this button will automatically be toggled to Off.
As a default, Background Discovery will not create pingable models for IP
addresses that could only be contacted via ICMP pings. If you want these
models to be created, toggle this menu button to the Yes setting.
IP Address List
Click this button to open the Show and Hide dialog box (Figure 2-5), which
lets you view and modify the list of IP addresses that Background Discovery
will attempt to contact.
Network Address
This field displays the network address of the model from which Background
Discovery was invoked. If the model was created manually with no network
address specified, this field will be blank, and the Status field (described
below) will display “No addresses to discover” when you attempt to start the
discovery session.
Subnet Mask
This field displays the subnet mask of the model from which Background
Discovery was invoked. If the model was created manually with no subnet
mask specified, this field will be blank, and the Status field (described below)
will display “No addresses to discover” when you attempt to start the
discovery session.
Thu May 07 14:42:03 1998
This field displays the starting date and time for the current or most recent
Background Discovery session for the selected model.
Thu May 07 14:42:48 1998
This field indicates when the most recent Background Discovery session for
the selected model ended. The field is blank if there is a session in progress.
This field displays the status message “Running” when a Background
Discovery session is in progress, or “No addresses to discover” if no addresses
9030727 E12
The User Interface
Background Discovery
The Show and Hide Dialog Box
remain on the Show and Hide dialog box’s Undiscovered Address list (in which
case, the Background Discovery button will be toggled to Off).
Community Names
If AutoDiscovery has already been run on behalf of the selected model, this
panel displays the community names used for the AutoDiscovery session. If
not, the default community name “public” is displayed. You can edit this panel
to add or delete names as needed for subsequent Background Discovery
sessions, but your edits will not affect a session that is already in progress.
Figure 2-5.
The Show and Hide Dialog Box
Show and Hide
Background Discovery ConÞguration
Undiscovered Address
Excluded Address
The Show and Hide Dialog Box
Accessible from the Background Discovery dialog box’s IP Address List
button, this dialog box lets you “show and hide” IP addresses in the sense that
you can determine which addresses will be included on the Undiscovered
Address list that Background Discovery will attempt to contact. Components
of the dialog box are discussed individually below.
The User Interface
AutoDiscovery User’s Guide
Background Discovery
The Show and Hide Dialog Box
Undiscovered Address
This scrollable list shows all the IP addresses that Background Discovery will
attempt to contact when next started for the selected model. As a default, the
list includes all in-range addresses for which models have not yet been
created, but you can also move addresses to this list from the Excluded
Address list using the arrow buttons described below. For a given session,
Background Discovery will attempt to contact each of the undiscovered
addresses once. New sessions will start automatically at the interval you have
specified until no addresses remain on this list.
Excluded Address
This panel lists all other in-range IP addresses that are not included in the
Undiscovered Address list. As a default, this list includes addresses for which
models have already been created, but you can also move addresses to this list
from the Undiscovered Address list using the arrow buttons described
<Clicking this arrow button moves a selected IP address from the Excluded
Address list to the Undiscovered Address list.
Clicking this arrow button moves a selected IP address from the
Undiscovered Address list to the Excluded Address list.
<<Clicking this arrow button moves all IP addresses from the Excluded
Address list to the Undiscovered Address list.
Clicking this arrow button moves all IP addresses from the Undiscovered
Address list to the Excluded Address list.
Available for each of the IP address lists in this dialog box, these menu
buttons provide two methods of locating a particular IP address or group of
addresses. When set to Filter, the list above the button will display only
addresses that contain whatever string you enter in the field to the right of
the button. When set to Search, the list will display only the first address
containing the string you have entered. You can then click the adjacent Next
button to display the next address containing that string.
9030727 E12
The User Interface
Background Discovery
The Show and Hide Dialog Box
This button is used only in conjunction with the Search option of the Search/
Filter button described above. Clicking Next displays the next IP address
that contains the string entered in the adjacent field.
Click this button to close the Show and Hide dialog box.
Click this button to cancel any changes you have maderedisplay the settings
that were in effect when the dialog box was opened.
The User Interface
AutoDiscovery User’s Guide
Chapter 3
Using AutoDiscovery
This chapter provides a step-by-step explanation of the recommended procedure for using
AutoDiscovery to create a comprehensive topological model of your network.
Before You Start
Although AutoDiscovery can be run on smaller networks that do not contain
routers or on network sections for which the routers are not accessible, the
initial discovery session for a larger network must be run at the Universe
level of the SPECTRUM topology scheme and there should be at least one
router model present. In fact, the more routers you have already modeled,
the better the coverage that will be achieved by your initial discovery session.
If you need to create a router model, perform the following steps:
1. Bring up SPECTRUM and navigate to the Universe-level Topology view.
2. Place the view in Edit mode and select New Model from the Edit menu.
This will display the Select Model Type dialog box, which lists the
available model types.
3. Select the appropriate router type by clicking on it, then click on OK to
display a model creation dialog box.
4. In the model creation dialog box, enter a valid model name, IP address
and, if applicable, an SNMP community name.
5. Click on OK. An icon representing the router model appears in the
Universe view. The icon displays a green label once SPECTRUM makes
contact with the device.
Creating the Initial Topological Model
The recommended procedure for using AutoDiscovery is to run it initially at
SPECTRUM’s Universe topology level with only Router Discovery enabled, so
9030727 E12
Creating the Initial Topological Model
that only Phase One discovery will be performed. As explained in Chapter2,
Phase One discovery models and places only routers, IP Class A, B, and C
networks, wide area links, and LANs. It does not discover details within the
discovered LANs. To have AutoDiscovery do Phase One discovery only,
perform the following steps:
1. Put the Universe View into Edit mode and select the AutoDiscover
option from the Edit menu.
2. When the main AutoDiscovery dialog box is displayed, ensure that only
the Router Discovery method is selected (the selection button next to the
method name appears in color if the method is selected). Deselect other
methods as necessary by clicking on them.
3. Use the IP Address Ranges panel to establish the boundaries of your
network by specifying at least one IP address range within which
AutoDiscovery will operate. For each range you wish to specify, enter the
low and high addresses in the IP Range field then click on the Add button.
See the discussion of IP Address Ranges in Chapter 2 for more
4. Look at the Discovery Protocols panel and make sure that SnmpPif is
selected. When the dialog box is initially opened, all four protocols are
selected. You can deselect any protocols that you know are not applicable
to your network; however, SnmpPif must be selected for Phase One
(router) discovery to operate properly.
5. If you wish to use an SNMP community name in addition to the default
(“public”), click in the Community Name field, then type in the name and
click on the Add button to add it to the list of community names. Repeat
as necessary for additional names, but remember that AutoDiscovery
searches your network for community names in the order they are listed
here. The first name a device responds to is the one that is used for the
model. If your network entails multiple community names and you have a
preference as to naming, be sure to list names in order of preference.
6. Click on the Options button (or select Options from the File menu) and
check the three optional settings. The “Hierarchy” mapping option should
be selected by default. Do not change the setting to “Flat” unless you want
all discovered elements to be placed within a single fanout model. Change
the Save Database setting to “True,” if desired. It is recommended that
you select one of the AutoPlace options (“Radial” or “Tree”) unless you
want to manually arrange the icons for discovered elements. Remember,
however, that in order for the AutoPlace feature to work, you must take
the topology view out of Edit mode before the discovery session is
7. Select the Settings option from the main dialog box’s File menu and then
select the Save Current option to write the current settings to the VNM
database. The saved settings will then become the default settings for
subsequent AutoDiscovery sessions.
Using AutoDiscovery
AutoDiscovery User’s Guide
Creating the Initial Topological Model
8. As noted in Step 6, take the topology view out of Edit mode if you want to
use AutoPlace, then click on the main dialog box’s Start button. This will
display the AutoDiscovery Status window. When the discovery session is
over, an “AutoDiscovery complete” message box will be displayed.
9. The Universe view should now contain one or more network icons (IP
Class A, B, or C) as well as icons for any routers connecting these
networks. (Note that the seed router(s) you started with may not remain
at this level; a router connecting subnets within an IP Class A, B, or C
network is placed within the Topology view for that network.) If you didn’t
select either the Radial or Tree AutoPlace options, put the Universe view
in Edit mode and use the mouse to drag the icons as necessary to arrange
them in whatever pattern you wish. If there are so many icons that the
view appears crowded, you can cluster network models as described at the
end of this chapter.
10. When the icons in the Universe view are arranged to your satisfaction,
navigate into the Topology views for each of the network models and
arrange the icons representing the LANs, internal routers, wide area
links, and other network entities discovered during Phase One. Here, too,
you may wish to cluster some of the subnets to avoid overcrowded views.
Be sure to follow the instructions in the section titled “Clustering
Networks” which appears at the end of this chapter.
Once you have arranged all the icons resulting from “router” discovery, the
next step is to expand your network model by running AutoDiscovery for each
of the LAN or other network models, this time selecting one or more of the
available discovery methods so that Phase Two (bridges) and Phase Three
(hubs) discovery will occur. The procedure is the same as that outlined above
for the Universe level with the following exceptions:
• Instead of starting AutoDiscovery from the Universe level, navigate to the
Topology view of the LAN or network model to be explored and select
AutoDiscovery from the view’s Edit menu.
• In the AutoDiscovery dialog box, make sure that at least one of the
discovery methods is selected, but do not use the Address Resolution Table
method by itself. Refer to Chapter2 for detailed descriptions of these
• The IP Address Ranges panel will show a default range that encompasses
all of the current LAN or network. If you wish to run AutoDiscovery for
only a portion of this range, delete the default and add the smaller range.
Note that any range specified at this level must be a subset of the range
you established previously at the Universe level.
Repeat the procedure for each of the other LAN or Network models, using the
appropriate methods, protocols, and SNMP community names.
9030727 E12
Using AutoDiscovery
Additional Discovery and Configuration Tips
Additional Discovery and Configuration Tips
As noted in Chapter 1, AutoDiscovery depends on the information available
from devices in your network at the time the application is run. If a device is
temporarily out of service, it will not be modeled. Also, if the information a
device contains has not been updated to reflect current configurations, then
AutoDiscovery may not be able to correctly model those configurations. For
these reasons, you may not discover every device in your network the first
time you go through the procedure described in the previous section, even
though you will have used all three discovery phases. Each additional
discovery session, however, will further refine your network model. Moreover,
subsequent sessions will take less time since you will have established
appropriate IP address ranges as defaults.
Another reason for running additional discovery sessions is to ensure correct
placement of routers if you opt to manually cluster discovered models as
described below.
Clustering Networks
If a discovery session generates a large number of IP Class A, B, or C network
icons at the Universe level, or a large number of LANs within an IP network
model, you may want to cluster groups of these specific network types within
generic network models (Model Type = Network). This will produce a less
cluttered-looking view and can make it easier to locate a given entity within
the overall network mapping scheme.
A cluster can be any meaningful grouping of IP networks or LANs. For
example, you might want to group all the networks associated with a specific
department in your organization, or all the LANs within a given building. The
only real prerequisite is that the organization scheme makes sense to you. To
divide a large network into more manageable clusters, do the following:
1. Navigate to the Universe-level Topology View and decide which IP
network icons you want to cluster.
2. Use the Edit menu’s New Icon option to create a model of type “Network”
for each cluster you’ve identified.
3. Use the Edit menu’s Erase, Cut, and Paste options to relocate each IP
network icon from the Universe level to within the Topology View for the
generic network model in which you want it clustered. Begin by erasing
any connection pipe icons that are attached to the IP network icon. Then
cut the IP network icon from the Universe View and paste it into the
Topology View for the generic network model.
Do NOT move any router or WA_Link models; these will be placed
correctly when you perform the next step.
Using AutoDiscovery
AutoDiscovery User’s Guide
Additional Discovery and Configuration Tips
Clustering Networks
4. Rerun AutoDiscovery at the Universe level. This will place routers and
wide area links within the correct clusters and will regenerate connection
pipe icons as necessary. Your clustering will be retained.
5. When discovery is complete, arrange the contents of the Universe and
generic network-level Topology Views to your preference.
6. Examine the Topology View for each of your IP network models. If there
are too many LAN models, cluster them within generic network models as
explained above, then rerun AutoDiscovery for that IP network and
arrange the resulting icons as desired. Again, do not move router icons;
allow AutoDiscovery to place them in the proper cluster.
When you have completed the procedures outlined in this chapter, you will
have an easily accessible, multilevel network model that is structured to allow
full exploitation of SPECTRUM’s power for effective monitoring and
management. You should continue to run AutoDiscovery periodically and after
known changes in your network’s configuration. See Appendix B for
instructions on running AutoDiscovery at predetermined intervals via the
UNIX “cron” facility or the NT Schedule Service.
9030727 E12
Using AutoDiscovery
Additional Discovery and Configuration Tips
Clustering Networks
Using AutoDiscovery
AutoDiscovery User’s Guide
Appendix A
IP Addresses
This appendix provides background on the structure and use of IP (Internet Protocol) addresses,
which are used to define the network boundaries within which AutoDiscovery will attempt to find
and model devices.
IP Address Structure
An IP address represents a connection to the network, not a hardware
interface. The address combines a network identifier segment (net ID) with a
host identifier segment (host ID). The net ID segment of an IP address is
assigned by the Internet Assigned Numbers Authority (IANA). The host ID
segment is determined by the network administrator.
IP addresses are expressed in “dotted quad” notation, i.e., they are made up of
four 8-bit parts, separated by “dots” (periods). Each of these binary parts is
written using decimal numbers and has a maximum decimal value of 255. The
function of each part differs depending on network class (A, B, or C) as
explained in the following sections.
Class A Networks
The most significant bit in the first part of the address is always 0. The Net ID
forms the first part and Host ID the last three parts.
Net IDs range
Host IDs range
Broadcast to all hosts in net
Full address range
0 to 127
0.0.0 to 255.255.254
255.255.255 (all ones) to
9030727 E12
IP Address Structure
Class B Networks
Class B Networks
The two most significant bits in first part of the address are 1 and 0
respectively. For class B, the Net ID forms the first two parts of address and
the Host ID the last two parts.
Net IDs range
Host IDs range
Broadcast to all hosts in net
Full address range
128.0 to 191.255
0.0 to 255.255
255.255 (all 1’s) to
Class C Networks
The two most significant bits in first part are 1 and 1. The Net ID forms the
first three parts of address and the Host ID only the last part.
Net IDs range from
Host IDs range
Broadcast to all hosts in net
Full address range is
192.0.0 to 255.255.255
0 to 254
255 (all 1’s) to
Deciphering an IP Address
Three steps are involved in reading an IP address. For example, to read the IP
address you would do the following:
1. Convert the first decimal segment (132) to its binary equivalent.
132 (decimal) = 1 0 0 0 0 1 0 0 (binary)
2. Next, identify the class of the network. In this case, it is a Class B
network because the first two bits of the first segment are 1 and 0,
3. Identify the host portion of the address. Since this is a Class B network,
the following must be true:
132.177 is the net ID, and 118.24 is the host ID.
IP Addresses
AutoDiscovery User’s Guide
Subnets and Subnet Addresses
Subnets and Subnet Addresses
A single IP network can be partitioned into multiple subnets by dividing the
host ID number into two parts: a subnet-number and a host-number. Figure
A-1 shows the breakdown of an IP address for a Class B network.
Figure A-1.
Class B Network IP Address
host ID
net ID
When referring to a subnet, and not any particular host in the subnet, the
host-number portion of the address is omitted or set to zero. For subnet
number 118 in the above example, the address to the full
For a Class C network, the first three parts of the IP address are used for the
net ID, leaving the fourth part as the host ID. To create a subnet number in
this case, you must break the 8-bits of the fourth part of the address into a
subnet-number and a host-number as shown in Figure A-2, where the net ID
is XXX.0.XX .
Figure A-2.
Class C Network IP Address
This scheme provides eight subnets (0 through 7), with 32 host-numbers for
each subnet (0 through 31), supporting a total of 256 addresses. Note that the
breakpoint between the subnet address and the host number is determined by
the system administrator. The subnet mask indicates which bits represent the
network or subnet address portion of the address and which represent the
host ID portion.
9030727 E12
IP Addresses
Subnets and Subnet Addresses
IP Addresses
AutoDiscovery User’s Guide
Appendix B
AutoDiscovery from the
Command Line
This appendix explains how to run AutoDiscovery from your operating system’s command line,
either by manual entry or by a crontab script that will automatically start the application at
regular predetermined intervals.
Entering Startup Commands
Once you have established your network bounds and other settings by
running AutoDiscovery through the user interface described in Chapters 2
and 3, you can start subsequent sessions from your operating system
command line by performing the following steps:
1. Start SpectroSERVER.
2. Navigate to the Install-Tools directory and enter one of the following
shell-specific commands:
source setup.csh (if you are in the c shell)
. setup.ksh (if you are in the korn or bourne shell)
3. Navigate to the SG-Tools directory.
4. Enter one of the following commands:
autodisc -vnm <machinename> -mh <modelhandle> -n -log
autodisc -vnm <machinename> -mh <modelhandle> -config
where the string machinename is the name of the machine on which
SpectroSERVER (the VNM) is running, and the string modelhandle is
the unique 6-digit hexadecimal number identifying the particular model
against which you want to run AutoDiscovery. To determine the
9030727 E12
Using crontab Scripts
modelhandle for a model in SPECTRUM, bring up the Command Line
Interface and enter the command: show models | grep <modelname>
(see the Command Line Interface User’s Guide for more information).
Note that if no modelhandle is specified, AutoDiscovery will run at the
Universe level. Use the -n argument if you wish to suppress display of the
AutoDiscovery dialog box. This is necessary in cases where the command
will be executed as part of a crontab script (see following section). When
AutoDiscovery’s user interface is suppressed, the -log parameter will
redirect all the progress and status messages normally displayed in the
AutoDiscovery Status window (see Chapter 2) to a logfile in the
AutoDisc.logs directory.
If you want the dialog box displayed so that you can reference and/or
modify previously established settings, you must use the Ðconfig
parameter instead. This will display the dialog box without the Start
button so that you cannot inadvertently start the session.
Using crontab Scripts
UNIX-derived operating systems allow you to define a process as a “cron job”
and run it automatically at predetermined intervals. To do this you must
create a crontab file that identifies a particular process as a cron job and
specifies the times when it will be run.
Since a windowing system is required to run AutoDiscovery and the
DISPLAY variable (normally set in your .profile file) must be set,
AutoDiscovery is not executed directly as a cron job from a crontab file.
Instead, the crontab file executes a script that first sets the environment for
AutoDiscovery, then calls it as a process of the script.
Therefore, to run AutoDiscovery as a cron job, you must first use a UNIX text
editor to write a simple script that sets the environment and provides the
executable name and any other applicable command line arguments. The
following is a sample script for a scenario in which SPECTRUM is installed in
the /usr/Spectrum directory and AutoDiscovery is located in the SG-Tools
# Run your .profile so that the DISPLAY variable is set
# Change to the SG-Tools directory, which contains
cd /usr/Spectrum/SG-Tools
# Run autodisc
AutoDiscovery from the Command Line
AutoDiscovery User’s Guide
Using crontab Scripts
autodisc -vnm wkstn1 -n
Note that the command that actually starts execution (autodisc -vnm
wkstn1 -n) is entered in the same format used when starting the application
directly from the command line as described earlier. You can create this script
anywhere you want; you tell the cron facility where to find it by the entry you
make in the crontab file as described below.
Crontab files are placed in the /var/spool/cron/crontabs directory.
Each line in a crontab file identifies a particular UNIX process and tells where
it resides and when it should be run. This information is specified in six fields
separated by spaces and arranged as follows:
minutes hours day-of-month month day-of-week command-sequence
1. minutes indicates number of minutes. Range is 0 to 59.
2. hours indicates number of hours. Range is 0 to 23.
3. day-of-month indicates day of month. Range is 1 to 31.
4. month indicates month of year. Range is 1 to 12.
5. day-of-week indicates day of week. Range is 0 to 6. Sunday = 0
6. command-sequence indicates the command sequence required to start a
process. Only the first line (up to a % character or end-of-line character) is
executed by the shell.
Fields 1 through 5, which tell UNIX when to start a process, can accept input
in any of the following formats:
a one or two-digit number, for example: 11
a series of numbers separated by commas, for example: 7,14,21
a range of numbers separated by a dash, for example: 1-7
an asterisk (*) character. The asterisk is either filler or a wild card. It
indicates that the job will be done for all possible values in a field, unless
obviated by another field; e.g., if you specify a certain day of the week, an
asterisk in the day-of-month field will be ignored.
Here are some examples of entries for fields 1 through 5. Each initiates a
process at the time and frequency indicated.
Daily at 10AM:
Daily at 3PM:
0 10 * * *
0 15 * * *
Weekly at 11:59AM on Friday:
59 11 * * 5
Weekly at 9AM on Monday, Tuesday, and Wednesday:
Twice monthly, on the 1st and 15th:
9030727 E12
0 9 * * 1-3
0 0 1,15 * *
AutoDiscovery from the Command Line
Using crontab Scripts
SPECTRUM Schedule Manager
In field 6, you specify the command sequence necessary to start a process. In
the case of AutoDiscovery, you simply provide the path to the directory where
the script is located and the name of the script.
Thus, to run the script you created to run AutoDiscovery as a cron job, do the
1. Create a crontab file using the following command:
crontab -e
This opens an existing crontab file (or creates one if none exists) using a
UNIX text editor.
2. Move to an empty line and enter the time specifications for the script
(fields 1-5), the path, and the script name. For example, to run the
AutoDiscovery script named adisc at noon every Friday, the entry would
appear as follows:
0 12 * * 5 /usr/myname/bin/adisc
3. Save your crontab file using the appropriate command for your platform
and editor. If your script and crontab file entry have been entered properly,
AutoDiscovery will run automatically at the specified time using the last
settings that were saved via the AutoDiscovery dialog box.
SPECTRUM Schedule Manager
Another way to run cron jobs on Solaris systems or to schedule automatic, periodic AutoDiscovery
sessions on Windows NT systems is through the SPECTRUM Control Panel’s Schedule Manager
feature. You can access this feature by clicking the Scheduler button in the Control Panel’s Server
Administration panel. The Schedule Manager provides a point-and-click interface for scheduling
execution of commands and scripts. For Solaris systems, Schedule Manager automatically
interprets your entries and places them in your crontab file. For Windows NT users, the Schedule
Manager provides access to NT Schedule Service. For more information, refer to About the
SPECTRUM Control Panel.
AutoDiscovery from the Command Line
AutoDiscovery User’s Guide
Appendix C
This appendix is a glossary of acronyms and technical terms used in this guide.
ARP Table: Address Resolution Protocol Table. Router table that resolves an
IP address to a physical (MAC) address. Read by AutoDiscovery to find a list
of devices with which the router has communicated.
Community Name: Used as a security element by the SNMP protocol.
Identifies the community to which a device belongs. Default name is “public.”
Connection Pipe: In DevTop, Cablewalk, and Topology views, logical
connections are represented by “pipe” icons. These connections do not
necessarily represent cables, but rather logical connections between devices.
crontab: A type of UNIX file that lists commands to be automatically
executed at specific dates or times. A crontab script can be used to initiate
AutoDiscovery sessions at predetermined intervals.
Domain Name: Part of naming hierarchy in an NIS environment. Identifies
a particular NIS name server.
FDDI: Fiber Distributed Data Interface – a high-speed, dual ring, token
passing network designed to run over multi-mode fiber optic cable.
ICMP: Internet Control Message Protocol. Part of Internet Protocol. Used to
handle errors and control messages at the IP layer.
Icon: A graphic symbol that represents a SPECTRUM model. Icons
incorporate double-click zones that provide access to views that display
operating statistics and configuration data for the model.
Inference Handler: A portion of the C++ code that provides intelligence to
the SpectroSERVER. An inference handler performs a single, specific task.
Interface Address: See IP Address.
9030727 E12
IP Address: A 32-bit number generally represented in dotted decimal
notation that is used to identify a specific host on a specific network. (See
Appendix A)
MAC_Address: See Physical Address.
MIB: Management Information Base. A collection of objects that can be
accessed via a network management protocol. Devices adhering to MIB
versions I and II are capable of being discovered by AutoDiscovery.
Model: A SpectroSERVER representation of a specific network device or
network group such as a specific bridge, cable, port, etc. See Model Type.
Model Handle: A 6-digit hexadecimal number that identifies a SPECTRUM
model. The model might be a Network Group, a LAN, or a Location such as a
building, room, or rack. This number appears in the title bar of each
Model Type: A template describing attributes, actions, and associations for
construction of a software model of a device. For example, the LAN_802_5
model type is used to create models of 802.5 LANs.
NFS: Network File Service. A distributed file system that allows a set of
computers to cooperatively access each other’s files in a transparent manner.
NIS: Network Information Service. A mechanism developed by Sun
Microsystems to consolidate the password, hosts, and ether files from many
UNIX workstations on to a few machines called servers. A server maintains
the files that describe hosts and users for a particular domain.
Null-Layer Suppression: The suppression of the display of intermediate
layers in a network hierarchy when these layers do not contain branching.
Octet / Octet String: An octet is an eight-bit word. An octet string is a series
of octets grouped together.
Off-Page Reference Icon: A topology icon for a device that is related to, but
not directly part of, the current view. The off-page icon usually implies a direct
connection to the current topology view. The model represented by an off-page
reference icon usually resides in a higher topological level.
Physical Address: A physical address is associated with a specific network
interface (such as an interface card). Replacing the network interface card
changes the physical address of that node. A physical address is six octets (48
bits). Physical addresses are generally assigned to a device at the time of its
Ping: Packet Internet Groper. A program that tests whether a destination is
reachable by sending an ICMP echo request and waiting for a reply.
AutoDiscovery User’s Guide
Protocol: A formal description of the messages to be exchanged and rules to
be followed, in order for two or more systems to exchange information.
Redundancy: A networking technique that reduces failures by assigning a
redundant communication path. Such paths are built into the network and a
means of enabling the redundant path is provided. Certain redundant paths
may not be detected by AutoDiscovery.
Relation: A relation is a classification describing how entities relate to each
other. Each relation contains a list of rules that apply the relation to model
types. Relations include (but are not restricted to): Encompasses, adjacent to,
Contains, Collects, HASPART, Connects to, and Monitors. See also Rules.
Rules: Applying a relation to specific model types creates a rule. For example,
in the Contains relation, “Network group contains LAN” is a rule stating that
a network group model can contain LAN models. See also Relation.
Subnet: A range of network addresses contained within a larger network.
Subnet Addressing: The splitting of the host portion of an Internet Address
into two sections. The left section describes a grouping of hosts as a subnet of
the IP network. The right-most section is used to assign a unique number to
each host within the subnet.
Subnet Mask: A bit mask used to select bits from an Internet address for
subnet addressing.
Segment: A media link connecting nodes in a network.
SNMP: Simple Network Management Protocol. A network management
protocol used in TCP/IP-based internets.
Universe: Highest level of SPECTRUM Topology views.
Unnumbered Link: A network segment that is not assigned an IP address.
Such links are not explicitly modeled by AutoDiscovery. Instead, they are
represented as discrete router-to-router connections.
Virtual Network Machine (VNM): The intelligent software in the
9030727 E12
AutoDiscovery User’s Guide
broadcast 2-5
destination 1-4
IP 2-4
next hop 1-4
physical (MAC) 2-4
subnet A-3
ARP Table Discovery 2-4
accessing 2-1
additional sessions 1-9, 3-4
benefits of using 1-1
constraints 1-9
definition 1-1
for larger networks 1-3
placement of discovered elements 2-9
saving settings 2-7
starting 2-7
from the command line B-1
status 2-9
stopping 2-7
time requirements 1-7
destination addresses 1-4
discovery methods 1-5, 2-3
Discovery Protocols 2-4
discovery protocols 2-5
Background Discovery 1-9, 2-1, 2-3, 2-5, 2-10
broadcast addresses 2-5
Clustering Networks 3-4
command line operation B-1
Community Names 2-11
community names 2-6, 2-14
creating a router model 3-1
cron job 1-9, 2-1, 2-5, B-2
CtronBrdgPif protocol 2-6
CtronIRMPif protocol 2-6
excluded addresses 2-15
hierarchical mapping 2-8
IcmpPif protocol 2-4, 2-6
IP address 2-4, 2-13
IP Address Ranges 2-4
IP Address Table 2-9
IP Routing Table 2-9
LAN Discovery 2-3
Live Pipes 1-3
logical connections 1-2
Lost and Found View 1-10
MAC address 2-4
Modeling and Mapping Conventions 1-2
classes of A-1
network address 2-13
next hop addresses 1-4
NIS Discovery 2-3
9030727 E12
Notice i
NT Schedule Service B-4
Phase One discovery 1-4, 2-3, 2-9, 3-2
Phase Three discovery 1-6, 2-4, 3-3
Phase Two discovery 1-5, 3-3
Phases of Discovery 1-3
physical address 2-4
Ping Retry Count 2-9
Ping Timeout 2-9
Pingables 2-11
pings 2-3
Progress Log 2-9
redundancy 1-9
redundant 1-9
Restricted Rights Notice ii
route tables 1-4
Router Discovery 2-3
rules for model/icon placement 1-2
seed routers 1-4
SNMP Community Names 2-6, 3-2
SnmpPif protocol 2-4, 2-6, 3-2
Solaris systems 2-3, B-4
spanning tree 1-10
SPECTRUM Schedule Manager B-4
subnet address A-3
subnet mask 2-13
undiscovered addresses 2-15
Virus Disclaimer i
Windows NT systems B-4
AutoDiscovery User’s Guide
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF