Guest

CiscoWorks Device Fault Manager

FAQs (1.x)

  • Viewing Options

  • PDF (559.6 KB)
  • Feedback
Device Fault Manager 1.x Frequently Asked Questions

Table Of Contents

Device Fault Manager 1.x Frequently Asked Questions

General Questions

DFM Domain Manager and Broker

Adding and Managing Devices

SNMP Traps

Polling

Device Discovery (Probing) and Inventory Collection

Faults and Alarms

Polling and Threshold Groups

Working with Interfaces

Adapters: Overview

File Notifier Adapter

Mail Notifier Adapter

Trap Notifier Adapter

HPOV-NetView Adapter

SNMP Trap Adapter

RME Adapter

Understanding and Modifying What DFM Console and Window Displays

Time and Timestamp Issues

Clearing and Deleting Alarms

Basic Installation Questions

Process Management


Device Fault Manager 1.x Frequently Asked Questions


General Questions
DFM Domain Manager and Broker
Adding and Managing Devices
SNMP Traps
Polling
Device Discovery (Probing) and Inventory Collection
Faults and Alarms
Polling and Threshold Groups
Working with Interfaces
Adapters: Overview
Understanding and Modifying What DFM Console and Window Displays
Basic Installation Questions
Process Management

General Questions

What does DFM do?
What is DFM not meant to be used for?
Does DFM maintain a history of faults?
How often is new device support released?

What does DFM do?

Identifies possible problems on Cisco devices by monitoring events sent by the device (traps) and querying the device for health (SNMP)

Interoperates with other fault management systems

Is ready to use out of the box, with preset threshold and polling parameters

What is DFM not meant to be used for?

MIB browsing/compiling

Trap viewing/listening (however, DFM does listen for and process certain SNMP traps)

Server/desktop fault analysis, or network-wide/multivendor fault analysis

Alerting users to device configuration changes

Does DFM maintain a history of faults?

DFM can maintain a history of faults if you download and install the Fault History dropin at http://www.cisco.com/cgi-bin/tablebuild.pl/cw-fault-history. You can also use the File Notifier Adapter to log all faults, and then, for example, write your own scripts for processing the fault file.

How often is new device support released?

DFM Incremental Device Support (IDU) packages are normally released every three months. These packages provide support for new devices, in between major DFM releases. (These packages normally also contain bug fixes.) You can download an IDU by logging into Cisco.com and pointing your browser to: http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm.

DFM Domain Manager and Broker

What is the DFM broker?
What is the DFM domain manager?
Can I change the name of the domain name to something besides DFM?
Can I define additional domains on the same system?

What is the DFM broker?

The DFM broker (called DfmBroker) is a process running DFM software that facilitates communication between a domain manager (DfmServer) and its clients (consoles and adapters). The broker maintains information about the domain manager: its name (DfmServer), the hostname on which the DfmServer is running, and the TCP port where the DfmServer listens for client connections. The broker periodically connects to DfmServer to monitor its status.

What is the DFM domain manager?

The DFM domain manager (called DfmServer) is a process running DFM software that monitors network elements, analyzes the causes of failures, and diagnoses the effects of the failures on related elements. The domain manager consists of the following:

DFM analysis model, which describes the elements DFM can manage, the faults managed elements can generate, and the symptoms caused by generated faults.

DFM inventory, an in-memory repository of elements managed by DFM.

DFM adapters, applications that perform special communications functions, such as updating inventory information, forwarding SNMP traps to another network management system (NMS), and receiving and forwarding fault information (a file, an email address, another NMS) to other recipients.

Can I change the name of the domain name to something besides DFM?

No.

Can I define additional domains on the same system?

No. Only one copy of DFM can be installed on a machine. This is the only supported configuration for more than one domain on the same machine.

Adding and Managing Devices

How do I add devices to DFM (so DFM can manage them)?
What is the difference between managed and unmanaged?
Are devices automatically managed when I add them to DFM?
Can I delete or unmanage multiple devices at one time?
Which devices can DFM manage?
How many devices can DFM manage?
How do I unmanage an element?
How do I add devices that have intelligent modules (such as a switch with an RSM or MSFC)?
What do I have to do if I change a managed device's community string?
What do I have to do if I change a managed device's IP address?
What determines whether or not a device element is managed?
How do I unmanage an element, or change an unmanaged element to managed?
How do I determine which ports and interfaces are managed or unmanaged?
What are the default managed/unmanaged ports/interfaces?
How do I change DFM's default behavior for determining whether or not to manage a device?

How do I add devices to DFM (so DFM can manage them)?

You can add devices to DFM in these ways:

Use a seed file to add a list of devices you want DFM to manage. A seed file is a two-column file listing devices by name/IP address and read-only community string.

For instructions on preparing a seed file, see the appropriate installation guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Be sure the default read community string matches that of your network.

The seed file must be on the same host as the domain manager. Once you have prepared your seed file, add the devices by selecting Inventory > Import from Seed File from the Administration Console. Be sure to select
Inventory > Reconfigure and Inventory > Save Inventory when you are done.

Use the Add Agent command to add one device. Select Inventory > Add Agent from the Administration Console. Be sure to select
Inventory > Reconfigure and Inventory > Save Inventory when you are done.

Use the RME Adapter to automatically synchronize a local or remote Resource Manager Essentials (Essentials) inventory with the DFM inventory. When devices are added to Essentials, they are automatically probed by DFM. (However, when devices are removed from Essentials, you must manually remove them from DFM.) The CiscoWorks process that performs this synchronization is called DfmChangeProbe.

What is the difference between managed and unmanaged?

Managed means that the element is monitored by the DfmServer. Unmanaged means that the element has been probed and element information is in the DFM inventory, but the DfmServer is not currently monitoring the element.

Are devices automatically managed when I add them to DFM?

Yes. DFM will attempt to manage any device it discovers during inventory collection. When a device is added, DFM probes it for configuration information, and adds that information to the DFM inventory. What goes into the inventory depends on the device MIBs, and to what extent DFM supports the MIB (or MIBs).

Can I delete or unmanage multiple devices at one time?

Yes, if you have installed the latest DFM IDU available from http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm.

Which devices can DFM manage?

DFM can manage all devices listed on the supported devices list (see http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/dev_sup/index.htm). Downloadable incremental device support (IDU) packages are available from http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm.

How many devices can DFM manage?

Standalone DFM can manage up to 1,000 devices and 30,000 ports (of which 15%, or 4,500, are managed ports). If DFM is installed with the LAN Management Solution (LMS) bundle, the bundle limit is 500 devices and the same number of ports (see http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/cw2000_b/lms/index.htm). You may occasionally want to verify how many ports are currently managed.

How do I unmanage an element?

From the Administration Console, select the element and right-click Unmanage. To unmanage a port or interface, select the element from the device's ComposedOf topology list, and right-click Unmanage. Be sure to select Inventory > Reconfigure and Inventory > Save Inventory when you are done. You can also unmanage all elements in a group (including a group of ports or interfaces) if you have installed the latest DFM IDU from http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm (see How do I unmanage an element?).

How do I add devices that have intelligent modules (such as a switch with an RSM or MSFC)?

Add the device before adding the module. Make sure that the community strings on the device and module are the same.

What do I have to do if I change a managed device's community string?

Change the device's community string in DFM by selecting Inventory > Add Agent from the Administration Console. Be sure you select
Inventory > Reconfigure and Inventory > Save Inventory when you are done.

What do I have to do if I change a managed device's IP address?

Do the following from the Administration Console:

1. Delete the device from the DFM inventory (select the device and right-click Delete).

2. Select Inventory > Reconfigure and Inventory > Save Inventory to save your changes to the inventory (which allow you to re-add the device).

3. Add the device to the DFM inventory using Inventory > Add Agent. (If your device is a switch that has intelligent modules, make sure the community strings for the switch and modules are the same.)

4. Select Inventory > Reconfigure and Inventory > Save Inventory.

What determines whether or not a device element is managed?

It depends on the settings in the polling and threshold groups the element belongs to. Of course, the device must also be supported by DFM (see http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/cw2000_b/lms/index.htm).

How do I unmanage an element, or change an unmanaged element to managed?

Users normally unmanage device elements they are not interested in. They also do so temporarily when they know there is an element problem and do not want to receive frequent notifications about it. Managing and unmanaging are performed from the Administration Console:

To unmanage an element, select the element and right-click Unmanage.

To manage an unmanaged element:

a. Select the element and right-click Manage.

b. Select the element and right-click Rediscover to probe the element.

c. Select Inventory > Reconfigure.

d. Select Inventory > Save Inventory.

To unmanage all elements in a group (including a group of ports or interfaces), use the new default threshold groups provided in the latest DFM IDU from http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm.

How do I determine which ports and interfaces are managed or unmanaged?

Use the sm_adapter command in conjunction with the getNetworkAdapters.asl script as described in (see the "Listing Managed Ports and Interfaces" section of the appropriate DFM user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/dfm123cd/u_gd/invent.htm). You must download and install DFM 1.2 Patch/IDU 1.2.10 or later to use this function.

What are the default managed/unmanaged ports/interfaces?

Ports (switches): By default, DFM manages trunk ports but does not manage access ports. DFM considers a port to be a trunk port if it connects to a Cisco network device running Cisco Discovery Protocol (CDP). In other words, a trunk port connects to a router, or to a switch that is managed by the same DFM server. DFM does not manage access ports by default (an access port is a switch port that is connected to a host or device not managed by DFM; that is, an end-station port).

Interfaces (routers): By default, DFM manages all interfaces listed in the ifTable.

How do I change DFM's default behavior for determining whether or not to manage a device?

To change this behavior, create a new polling and threshold group for the elements you want to be managed by default.

SNMP Traps

Does DFM report all SNMP traps?
How does DFM handle SNMP traps?
What version of SNMP traps does DFM support?
How do I configure DFM to receive traps?
How do I configure DFM to forward traps?

Does DFM report all SNMP traps?

No, because DFM is not a trap viewer. However, certain traps are processed (see the MIBs/SNMP traps appendix in the appropriate DFM user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

How does DFM handle SNMP traps?

For certain SNMP traps, DFM will either:

Process the SNMP traps, or

Treat the SNMP traps as pass-through traps and display them on the Monitoring Console.

For more information, see the MIBs/SNMP trap appendix in the appropriate DFM User Guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

What version of SNMP traps does DFM support?

DFM 1.0 and 1.1 pollers support V1 SNMP traps for polling, receiving and forwarding. The DFM 1.1 poller can support V2 traps if you download and install the new Incremental Device Update (IDU) packages, available from Cisco.com at http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm. Received V2 traps are forwarded as V1 traps.

DFM 1.2 and DFM 1.2 Updated for Common Services Version 2.2 support V1 and V2 traps for polling and receiving. Received V2 traps are forwarded as V1 traps.

How do I configure DFM to receive traps?

DFM is already configured to receive SNMP traps, but you can specify a different listening port with the SNMP Trap Adapter.

How do I configure DFM to forward traps?

Use the SNMP Trap Adapter.

Polling

What is polling?
Does DFM use only SNMP for polling?
What MIB objects does DFM poll?
By default, how often is polling done?
Can I use DFM to browse MIBs?
Do I need to download any MIBs from Cisco.com to use DFM with another NMS (such as HP OpenView)?
Where is the DFM-MIB?

What is polling?

Polling is the process of requesting and receiving information from a particular device MIB (such as the CISCO-ENVMON MIB for voltage/temperature faults), or a particular MIB object. Polling does not reveal configuration changes, such as the addition of new interfaces. What is polled is determined by the polling and threshold groups to which a device element belongs.

Does DFM use only SNMP for polling?

No. DFM collects fault and performance information using SNMP, but device connectivity is monitored using ICMP (ping). If a device does not respond to an ICMP poll, it is placed on a "do not poll" list and is not polled by SNMP. In addition, the polling interval for ICMP is always set to be 60 seconds less than the pollling interval for SNMP, to ensure that SNMP has the most current and correct data. (You can disable ICMP polling without disabling SNMP polling if you have downloaded and installed the latest patch/IDU from the DFM download site: http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm.)

What MIB objects does DFM poll?

The MIBs that DFM polls are listed in the MIBs/SNMP traps appendix in the appropriate user guide (see http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm). The DFM patch/IDUs often add additional MIB support.

By default, how often is polling done?

SNMP polling is done every four minutes, by default. An explanation of how the polling interval is calculated is provided in the defaults chapter in the appropriate user guide (see http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

Can I use DFM to browse MIBs?

No. DFM is not a MIB browser.

Do I need to download any MIBs from Cisco.com to use DFM with another NMS (such as HP OpenView)?

No.

Where is the DFM-MIB?

Because it is not a CISCO MIB, the DFM-MIB is not available on any download site, although the contents are publicized in an appendix in the appropriate user guide (see http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

Device Discovery (Probing) and Inventory Collection

What is probing?
Will DFM detect changes I make to my devices/network configuration?
Can I view the inventory file of managed devices?
How often is the DFM inventory saved?
Does DFM automatically discover new devices in the network?
Does DFM automatically rediscover existing devices in the network?
What is the difference between discovery, rediscovery, and inventory collection?
Can I change the inventory collection schedule?
What is the difference between using the Rediscovery Schedule and the Automatic Inventory Collection checkbox?
How do I disable the Automatic Inventory Collection checkbox?
Can I perform immediate manual discover/rediscovery?
What is device certification?
Do I have to wait for another DFM release for new device support?

What is probing?

Probing is done during rediscovery: it is the process of requesting and receiving device configuration information by performing a complete device MIB walk (which is how DFM discovers which MIBs a device supports). This is done when a new device is added to the DFM inventory (and the device is discovered), when inventory collection is done and devices are rediscovered, or when you perform a manual rediscovery of a device element. The device configuration DFM receives during a probe is saved to the DFM inventory.

Will DFM detect changes I make to my devices/network configuration?

Yes, but only when inventory collection is done and the device is rediscovered. You can rediscover a device manually by selecting the device and right-clicking Rediscover. (Be sure to select Inventory > Reconfigure and
Inventory > Save Inventory after a manual rediscovery.)

Can I view the inventory file of managed devices?

No. The DFM inventory is a runtime, in-memory inventory, and is not viewable. There is a permanent inventory that is saved to disk, but it is not viewable. All changes you make are stored in the runtime inventory when you select
Inventory > Reconfigure, but changes are saved to the permanent inventory only when inventory collection is done or when you select
Inventory > Save Inventory. (Remember that, in addition to device information, the DFM inventory contains information on the relationships among device objects, polling and threshold groups and their settings, and so forth.)

How often is the DFM inventory saved?

The inventory is saved every six hours (but inventory collection is done once a week).

Does DFM automatically discover new devices in the network?

No; devices must be added using one of the supported methods.

Does DFM automatically rediscover existing devices in the network?

Yes. By default, this is scheduled to be done once a week (on Sunday at midnight), and is called inventory collection. Devices are rediscovered for configuration information, and based on that collection, DFM updates the DFM inventory.

What is the difference between discovery, rediscovery, and inventory collection?

These concepts are basically the same. Inventory collection occurs when DFM probes the inventory to discover new devices, or to rediscover currently managed devices. DFM updates the inventory based on the results of inventory collection. A complete inventory collection is performed, by default, once a week (on Sunday at midnight).

Can I change the inventory collection schedule?

Yes, by using the Rediscovery Schedule. Select Device Fault Manager > Administration > Device Discovery  > Rediscovery Schedule, click Help, and follow the instructions.

What is the difference between using the Rediscovery Schedule and the Automatic Inventory Collection checkbox?

The Automatic Inventory Collection checkbox (which is displayed when you select the DFM domain in the topology window) runs on a relative time basis. In other words, it runs relative to the last collection. The Rediscovery Schedule, on the other hand, allows you to specify a date, time, and period for collection. Because the Rediscovery Schedule function is consistent with how other CiscoWorks functions work, we recommend that you use the Rediscovery Schedule and disable the Automatic Inventory Collection checkbox.

How do I disable the Automatic Inventory Collection checkbox?

From the Administration Console:

1. Select the DFM domain in the topology window.

2. Deselect the Automatic Inventory Collection checkbox.

3. Click Apply, and select Inventory > Reconfigure.

4. Select Inventory > Save Inventory.

Can I perform immediate manual discover/rediscovery?

Yes, you can probe in one of three different ways, all from the Administration Console:

To probe only a selected device, select Inventory > Rediscover.

To probe everything in the update pending list, select
Inventory > Inventory Collection Pending.

To probe the entire inventory, select Inventory > Inventory Collect All.

What is device certification?

Device certification describes the level at which DFM supports a device. The higher the level, the more information DFM can provide about the device. You can check a device's certification level by selecting the device, right-clicking Browse, and checking the value of the Certification attribute in the Attributes tab. There are five levels of DFM device certification:

Validated—DFM has been field-tested with this device. (This is the highest level of certification.)

Certified—DFM uses information public MIBs to perform discovery on this device.

Template—DFM has no information on MIBs supported by this device.

Undiscovered—DFM has not fully discovered the device due to an inventory collection error (for more details, see the DFM inventory chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

Unsupported—DFM does not support this device.

Devices must be either Validated or Certified to be considered supported by DFM. Note that the Undiscovered level indicates an inventory collection error, not a lack of device support.

For more information on device certification levels, see the collection chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Do I have to wait for another DFM release for new device support?

No. Between major releases, support for certain Cisco devices can be downloaded. These Incremental Device Update (IDU) packages are available to all users from Cisco.com at http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm.

Faults and Alarms

What is a fault, and how does DFM classify faults?
What is the difference between the DFM terms fault and alarm?
What are the most common faults and exceptions diagnosed by DFM?
What determines how DFM performs polling and generates alarms for devices?

What is a fault, and how does DFM classify faults?

DFM considers an event to be a fault when it surpasses a threshold for acceptable behavior. When this happens, DFM generates a fault notification (or alarm), which is displayed in the Monitoring and Administration Consoles. DFM displays two kinds of fault notifications:

Symptomatic notifications, which identify a single fault (symptom).

Compound notifications, which identify one or more related symptoms. In DFM terminology, a compound notification is a roll-up of related symptoms.

Compounds are sometimes called aggregates (in the adapter configuration files), or simply exceptions (from the naming standard for compounds—OperationalException, PerformanceException, and so forth).

What is the difference between the DFM terms fault and alarm?

A fault is an event in which DFM determines a device is operating outside acceptable behavior—for example, a port is down that should be up, or a threshold has been exceeded that should not be exceeded. An alarm is a notification created by DFM to inform the user about the fault. DFM gathers information about the device and the fault, and then issues the alarm.

What are the most common faults and exceptions diagnosed by DFM?

The following are examples of the what DFM commonly looks for on different types of elements:

Chassis—Backplane utilization

Fan—Fan state not normal

Memory—Excessive fragmentation, buffer miss rate, buffer utilization, free memory

Network adapters—Backup activated error rates (at system and VLAN level), broadcast rates, collision rates, discard rates, flapping, maximum uptime, queue drop rates, utilization

Power supplies—Voltage out of range

Processors—Utilization

SNMP agent—Unresponsive

System—Excessive restarts

Temperature—Temperature out of range

For more details on faults reported by DFM, see the fault chapter in the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

What determines how DFM performs polling and generates alarms for devices?

This is determined by the polling and threshold groups to which a device and its elements belong.

Polling and Threshold Groups

What determines the defaults for DFM polling and alarm generation?
When a device is added to DFM, how does DFM determine which polling and threshold group the device should belong to?
What are classes and instances?
What are the default polling and threshold groups?
Can an element belong to more than one polling group or threshold group?
Can I create new groups?
How do I disable polling for a device or specific elements in a device?
How do I disable only certain types of polling?
How do I change polling for unwanted faults/exceptions?
How do I change thresholds for unwanted faults/exceptions?

What determines the defaults for DFM polling and alarm generation?

This is determined by the polling group and threshold group to which a device and its elements belong. When a device is added to DFM and has been discovered, the elements of the device become members of two groups—a polling group and a threshold group. These groups determine management policies for the device. The polling group determines which device elements are polled and how often; the threshold group determines at which point DFM should generate alarms for an element. There are different types of polling groups and threshold groups, and you can create your own groups.

When a device is added to DFM, how does DFM determine which polling and threshold group the device should belong to?

The device is assigned to polling and threshold groups depending on the group matching criteria as described in the DFM user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

What are classes and instances?

A class is a group of objects that are structurally and operationally related, such as switches or routers. An object or element that belongs to a class is called an instance of that class; for example, a specific Cisco Catalyst 550 device is an instance of the Switch class. The most commonly used DFM classes are described in an appendix in the appropriate DFM user guide (see http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

The terms object and are also used interchangeably, although an element normally specifies something physical (such as a power supply), while an object can specify something logical (such as a VLAN).

What are the default polling and threshold groups?

The following are the default groups:

Default polling groups: Switches, Routers, Hubs and Bridges, Uncertified Systems, and Other Systems. For more information on the default polling groups, see the group chapter in the appropriate DFM user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Default threshold groups: Interface, Access Ports, Trunk Ports, System Resource. The latest DFM patch/IDU added three ISDN groups to the default interface groups. (By default, access ports are not managed, which is why that group will be empty.) If you have downloaded and installed Patch/IDU 1.2.9 or later, there are three additional threshold groups: System Elements, Unmanaged Ports/Interfaces, and Unmanaged systems (go to http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm.) For more information on the default threshold groups, see the groups chapter in the appropriate DFM user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

Can an element belong to more than one polling group or threshold group?

Yes. If an element belongs to more than one group, DFM will use the settings of the group with the highest priority For more information, see the groups chapter in the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Can I create new groups?

Yes. The following are examples of why you might want to do this:

To poll certain routers more frequently

To change DFM default behavior regarding unmanaged elements (so they are managed by default)

To set port flapping thresholds on specific router interfaces

The user guide explains how to create groups (see the groups chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm). If you want to create a group that is a subset of an existing group—such as a group of routers—select the existing group, and then select Group > New Group. You can then adjust the settings as desired. Be sure to select Apply, Inventory > Reconfigure and Inventory > Save Inventory when you are done.

How do I disable polling for a device or specific elements in a device?

You should unmanage the device if you want to disable polling for all elements in a device, or if you want to disable polling for a group of devices. To disable polling for only certain elements in a device (such as ports), select the specific element and right-click Unmanage. (Be sure to select Inventory > Reconfigure and Inventory > Save Inventory when you are done.)

You can also disable ICMP/connectivity polling (without disabling SNMP/availability polling) if you have installed the latest DFM IDU (from http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm).

How do I disable only certain types of polling?

Disable the AnalysisMode setting for that type of polling. For example, to poll ports and interfaces for performance but not for connectivity, do the following:

1. From the Polling and Thresholds Console, select the Polling tab and locate the polling group you want to reconfigure.

2. Under Settings, select the type of polling you want to disable (for example, Connectivity Polling—Ports and Interfaces).

3. In the right window, set AnalysisMode to Disable.

4. Click Apply.

5. Select Group > Reconfigure.

6. To permanently save your changes, from the Administration Console, select Inventory > Save Inventory.

You can also disable ICMP/connectivity polling (without disabling SNMP/availability polling) if you have installed the latest DFM IDU (from http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm).

How do I change polling for unwanted faults/exceptions?

From the Polling and Thresholds Console, select the Polling tab. Select the polling type on the left, and adjust the threshold on the right. Be sure to do the following to save your settings:

1. Click Apply.

2. Select Group > Reconfigure.

3. From the Administration console, select Inventory > Save Inventory.

How do I change thresholds for unwanted faults/exceptions?

From the Polling and Thresholds Console, select the Thresholds tab. Select the setting on the left, and adjust the threshold on the right. Be sure to do the following to save your settings:

1. Click Apply.

2. Select Group > Reconfigure.

3. From the Administration Console, select Inventory > Save Inventory.

Working with Interfaces

How does DFM determine an interface mode?
How do I change an interface mode?
If I reload a device, will DFM automatically recognize the device's ifIndex change?
How does DFM handle port/interface duplexity?
How do I disable management for virtual interfaces?

How does DFM determine an interface mode?

DFM classifies most interfaces as NORMAL, with the following exceptions:

PPP, SLIP—Dial-on-Demand

BRI, ISDN—Backup

(ISDN interfaces are modeled differently if you have installed the latest DFM IDU available from http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm.)

How do I change an interface mode?

Although it appears as if you can edit the group attribute fields using the Administration Console, such changes do not affect the settings. You must use the Polling and Thresholds Console and create a new group to change the interface mode. The following tells you how to change PPP interfaces from the default mode (Dial-on-Demand) to NORMAL:

1. From the Polling and Thresholds Console, in the Thresholds tab, create a new interface group called PPP.

2. Add the following Matching Criteria: TYPE=PPP, MODE=NORMAL

3. Increase the Priority of this new Multilink PPP Group to a higher level than the default group (Dial-on-Demand).

4. Click Apply.

5. Click Reconfigure.

6. Verify the members of the new PPP group.

7. From the Administration Console, select Inventory > Save Inventory.

If I reload a device, will DFM automatically recognize the device's ifIndex change?

No, not until inventory collection is done. You should manually rediscover the device from the Administration Console:

1. Select the device and right-click Rediscover.

2. Select Inventory > Reconfigure.

3. Select Inventory > Save Inventory.

How does DFM handle port/interface duplexity?

DFM determines duplexity in different ways, depending upon which version of DFM you are running. For details see the information on the high utilization fault (in the faults chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm. If you know that the value is incorrect, you can change it manually (see the groups chapter in the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.)

How do I disable management for virtual interfaces?

If you have downloaded and installed DFM 1.2 Patch/IDU 1.2.9 (or later) you can perform bulk manage and unmanage operations on selected groups of ports or interfaces.

1. From the Polling and Thresholds Console, in the Thresholds tab, select the port or interface group you wish to unmanage.

2. Right-click Unmanage.

3. Select Reconfigure from the Group menu.

4. Select Save Inventory from the Inventory menu to update the DFM inventory.

If you have not downloaded and installed the latest patch/IDU, to disable analysis for interfaces (or ports), assign the interfaces to a new group that has no settings:

1. From the Administration Console, select Edit > Polling and Thresholds.

2. In the Thresholds tab, expand the topology tree until you have expanded Interface Groups.

3. Select Interface Groups and right-click New Group.

4. Enter a new name for the group (for example, "Virtual Interfaces") and an optional description. The new group is displayed under Interface Groups.

5. Select the new group and the Matching Criteria tab.

6. Set the matching criteria:

a. Select DeviceID and/or Description from the Disabled Criteria list.

b. Click Enable.

c. Double-click the Value column to change the value used in matching criteria (this example is for virtual interfaces):

For DeviceID, use "*VL*"

For Description, use "*VLAN*"

d. Click Register and Apply.

7. Under the Priorities tab, select the new group you created.

8. Using the arrow button, move the new group to the top of the list (to make it highest priority).

9. Click Apply.

10. Select the DFM domain and Group > Reconfigure.

11. From the Administration Console, select the DFM domain and
Inventory > Save Inventory.

Adapters: Overview

What are DFM adapters?
What are the basics of adapter configuration?
How do I set up notification adapters to monitor only selected faults?
What are the names of the CiscoWorks adapter processes?
File Notifier Adapter
Mail Notifier Adapter
Trap Notifier Adapter
HPOV-NetView Adapter
SNMP Trap Adapter
RME Adapter

What are DFM adapters?

Adapters are applications that run as separate processes connecting the DfmServer to its environment. Adapters allow DFM to communicate and exchange information with other applications. DFM adapters can be classified as follows:

Notification adapters, which forward event information to designated recipients:

A file (File Notifier Adapter)

An email address (Mail Notifier Adapter)

Another NMS (Trap Notifier Adapter)

Event adapters, which forward traps from another NMS to DFM. The HPOV-NetView Adapter performs this function for HP OpenView and NetView.

Special adapters, which perform special functions:

SNMP Trap Adapter, which receives and/or forwards pass-through and certain processed SNMP traps.

RME Adapter, which synchronizes the DFM inventory with the list of devices in the Resource Manager Essentials inventory.

What are the basics of adapter configuration?

The following are adapter starting points:

You do not have to do anything to start these adapters; they start automatically when installed:

The RME Adapter starts automatically if Resource Manager Essentials (Essentials) is installed on the same machine as the adapter. Even if Essentials is installed after the adapter, the adapter detects the installation and starts automatically (unless it is manually disabled).

The HPOV-NetView Adapter starts automatically if HP OpenView or NetView is installed on the same machine. If HP OpenView or NetView is installed after the adapter, the adapter detects the installation and starts automatically.

The SNMP Trap Adapter (for trap receiving).

You must manually start this adapter, but you do not need to configure it unless you want to use a listening port other than 162:

File Notifier Adapter

You must configure and manually start these adapters:

Mail Notifier Adapter

Trap Notifier Adapter

SNMP Trap Adapter (for forwarding)

If the CiscoWorks daemon manager goes down, adapters requiring manual starts will not be restarted. You can use the command line to modify this behavior and register these adapters for automatic restart.

How do I set up notification adapters to monitor only selected faults?

You must manually edit the adapter configuration file, and then restart the appropriate CiscoWorks adapter process.

You can do one of the following:

Specify an existing subscription profile (filename) by using this code fragment:

SubscribesTo =
{
	GA_ProfileSubscription::Descriptive-Text
	{
		profileName = "filename"
	}
}

For more information, see the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Specify the classes, instances, and events you want the notification adapter to track by using this code fragment (do not use a comma after the final fragment):        

	GA_ChoiceSubscription::Descriptive-Text
	{
	# Subscribe to events whose class, instance, and event
	# names match the given pattern.
		 className = "class"
	      instanceName = ".*"
		 eventName = "event"
		aggregates = {TRUE | FALSE}
		  symptoms = {TRUE | FALSE}
	},
	GA_ChoiceSubscription::Descriptive-Text
	{
	# Subscribe to events whose class, instance, and event
	# names match the given pattern.
		 className = "class"
	      instanceName = ".*"
		 eventName = "event"
		aggregates = {TRUE | FALSE}
		  symptoms = {TRUE | FALSE}
	}

Note that you cannot use include or exclude patterns, and you can use the wildcard pattern only for instanceName( for subscription examples, see the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

What are the names of the CiscoWorks adapter processes?

The following adapters are CiscoWorks processes:

File Notifier Adapter: DfmFileNotifier

Mail Notifier Adapter: DfmMailNotifier

Trap Notifier Adapter: DfmTrapNotifier

RME Adapter: DfmChangeProbe

SNMP Trap Adapter: part of DfmServer process

The HPOV-NetView Adapter does not have a corresponding CiscoWorks process.

File Notifier Adapter

What does the File Notifier Adapter do?
Can I modify the content of the information saved by the File Notifier Adapter?
Where is the File Notifier Adapter configuration file located?
Where is the File Notifier Adapter log file?
Where is the log of alarms stored by the File Notifier Adapter?
What do I have to do to get the File Notifier Adapter running?
How do I specify which alarms I want logged by the File Notifier Adapter?
What do I have to do if I make a manual change to the File Notifier Adapter configuration file?
How do I disable the File Notifier Adapter?

What does the File Notifier Adapter do?

The File Notifier Adapter creates an alarm log of all notifications processed by DFM, and stores the notifications in a file. A file is the only valid recipient for this adapter, and you cannot change the location of the alarm log file (NMSROOT/objects/smarts/logs/DFM-alarms.log).

Can I modify the content of the information saved by the File Notifier Adapter?

You can modify the types of faults that are logged, but you cannot change the format of the file. The following is an example of the file contents:

02-Feb-2001 11:58:38 NOTIFY Switch::172.16.0.0::DiscoveryError 100% An 
error was encountered during the last discovery probe of this System.  

02-Feb-2001 12:00:19 CLEAR Switch::172.16.0.0::DiscoveryError An error 
was encountered during the last discovery probe of this System.  

02-Feb-2001 12:03:14 NOTIFY Switch::172.16.0.0::PowerSupplyException 
100% System is experiencing power supply problems.  

Where is the File Notifier Adapter configuration file located?

The File Notifier Adapter configuration file is at NMSROOT/objects/smarts/conf/notifier/filelog_notify.conf.

Where is the File Notifier Adapter log file?

The File Notifier Adapter log file is at NMSROOT/objects/smarts/logs/sm_file_notifier.log.

Where is the log of alarms stored by the File Notifier Adapter?

The File Notifier Adapter alarm log file is at NMSROOT/objects/smarts/logs/DFM-alarms.log.

What do I have to do to get the File Notifier Adapter running?

Select Device Fault Manager > Administration > Fault Notification > File Notifier, click Help, and follow the instructions.


Note DFM 1.0 does not offer this function from the GUI. To use this adapter, you must start the adapter DfmFileNotifier process using Server Configuration > Administration > Process Management > Start Process.


To log only certain alarms, you must edit the configuration file. If the CiscoWorks daemon manager goes down, the adapter will not be restarted when DFM comes up. If you want it to automatically restart, you must register the adapter as described in the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

How do I specify which alarms I want logged by the File Notifier Adapter?

You must manually edit the configuration file to subscribe to specific events.

What do I have to do if I make a manual change to the File Notifier Adapter configuration file?

You must restart the DfmFileNotifier process for your changes to take effect.

How do I disable the File Notifier Adapter?

Select Device Fault Manager > Administration > Fault Notification > File Notifier, click Help, and follow the instructions.


Note DFM 1.0 does not offer this function from the GUI. To disable this adapter, you must stop the adapter DfmFileNotifier process using Server Configuration > Administration > Process Management > Stop Process.


If you have configured the adapter to restart upon CiscoWorks restart and you want to disable automatic restart of the adapter, see the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Mail Notifier Adapter

What does the Mail Notifier Adapter do?
Do I have to have an SMTP server to use the Mail Notifier Adapter?
Where is the Mail Notifier Adapter configuration file?
Where is the Mail Notifier Adapter log file?
What do I have to do to get the Mail Notifier Adapter running?
Can I use the Mail Notifier Adapter to send a page when a certain fault occurs?
Can I modify the content of the email sent by the Mail Notifier Adapter?
Can I configure the Mail Notifier Adapter to send notifications to different recipients depending on the fault type?
What is the difference between using the Mail Notifier Adapter and Edit > Recipients in the Monitoring Console?
Can I configure the adapter to send email only when certain alarms are generated?
What do I have to do if I make a manual change to the Mail Notifier Adapter configuration file?
How do I disable the Mail Notifier Adapter?

What does the Mail Notifier Adapter do?

The Mail Notifier Adapter sends email to an email address when DFM generates certain notifications.

Do I have to have an SMTP server to use the Mail Notifier Adapter?

Yes.

Where is the Mail Notifier Adapter configuration file?

The Mail Notifier Adapter configuration file is at NMSROOT/objects/smarts/conf/notifier/mail_notify.conf.

Where is the Mail Notifier Adapter log file?

The Mail Notifier Adapter log file is at NMSROOT/objects/smarts/logs/sm_mail_notifier.log.

What do I have to do to get the Mail Notifier Adapter running?

Select Device Fault Manager > Administration > Fault Notification > Mail Notifier, click Help, and follow the instructions.


Note DFM 1.0 does not offer this function from the GUI. To use this adapter, you must start the adapter DfmMailNotifier process using Server Configuration >Administration> Process Management > Start Process.


To log only certain alarms, you must edit the configuration file. If the CiscoWorks daemon manager goes down, the adapter will not be restarted when DFM comes up. If you want it to automatically restart, you must register the adapter as described in the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Can I use the Mail Notifier Adapter to send a page when a certain fault occurs?

Yes, if the email address you provide is a paging gateway.

Can I modify the content of the email sent by the Mail Notifier Adapter?

You can modify the types of faults that are logged, but you cannot modify the content of the email. The following is an example of an email sent by the Mail Notifier Adapter:

InCharge Domain: DFM Router 10.3.28.1 OperationalException - This 
system or the components of this system are not functioning properly

Can I configure the Mail Notifier Adapter to send notifications to different recipients depending on the fault type?

No. You can send notifications to different recipients, but all recipients will receive the same information.

What is the difference between using the Mail Notifier Adapter and
Edit > Recipients in the Monitoring Console?

The Edit > Recipients function is not supported. Use Device Fault Manager > Administration > Fault Notification > Mail Notifier.

Can I configure the adapter to send email only when certain alarms are generated?

Yes. You must manually edit the configuration file to subscribe to specific events.

What do I have to do if I make a manual change to the Mail Notifier Adapter configuration file?

You must restart the DfmMailNotifier process for your changes to take effect.

How do I disable the Mail Notifier Adapter?

Select Device Fault Manager > Administration > Fault Notification > Mail Notifier, click Help, and follow the instructions.


Note DFM 1.0 does not offer this function from the GUI. To disable this adapter, you must stop the adapter DfmMailNotifier process using Server Configuration > Administration > Process Management > Stop Process.


If you have configured the adapter to restart upon CiscoWorks restart and you want to disable automatic restart of the adapter, see the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Trap Notifier Adapter

What does the Trap Notifier Adapter do?
Where is the Trap Notifier Adapter configuration file?
Where is the Trap Notifier Adapter log file?
Can I forward traps to multiple NMSs using the Trap Notifier Adapter?
Can I configure the Trap Notifier Adapter to sent notifications only when certain alarms are generated?
Can I configure the Trap Notifier Adapter to send notifications to different NMSs, depending on the fault type?
What do I have to do to get the Trap Notifier Adapter running?
What do I have to do if I make a manual change to the Trap Notifier Adapter configuration file?
How do I disable the Trap Notifier Adapter?

What does the Trap Notifier Adapter do?

The Trap Notifier Adapter converts DFM alarms displayed on the Monitoring Console into SNMP traps, and forwards these traps to an NMS, normally for additional processing or display. The converted alarms can also be forwarded to another domain manager. DFM has its own MIB for generating the SNMP traps; you do not have to download the MIB to use it. (Although the MIB is not on a download site, you can view its contents in the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.)

Where is the Trap Notifier Adapter configuration file?

The Trap Notifier Adapter configuration file is at NMSROOT/objects/smarts/conf/notifier/trap_notify.conf.

Where is the Trap Notifier Adapter log file?

The Trap Notifier Adapter log file is at NMSROOT/objects/smarts/log/sm_trap_notifier.log.

Can I forward traps to multiple NMSs using the Trap Notifier Adapter?

Yes, but you will have to manually edit the configuration file and restart the CiscoWorks DfmTrapNotifier process. The following excerpt from the configuration file shows the portion of the file that you must edit:

#For case of three recipients:

ProvidesAdditionalParams = 
Trap_AdapterParams::trap_Notifier-Parameters
     {
        recipients = {{"host_name1", 162, "1"},
                       {"host_name2", port_num2, "1"},
                        {"host_name3", port_num3, "1"} }
     }

Can I configure the Trap Notifier Adapter to sent notifications only when certain alarms are generated?

Yes. You must manually edit the configuration file to subscribe to specific events.

Can I configure the Trap Notifier Adapter to send notifications to different NMSs, depending on the fault type?

No. You can send traps to different NMSs, but all NMSs will receive the same information.

What do I have to do to get the Trap Notifier Adapter running?

Select Device Fault Manager > Administration > Fault Notification > Trap Notifier, click Help, and follow the instructions.


Note DFM 1.0 does not offer this function from the GUI. To use this adapter, you must start the adapter DfmTrapNotifier process using Server Configuration > Administration > Process Management > Start Process.


To log only certain alarms, you must edit the configuration file. If the CiscoWorks daemon manager goes down, the adapter will not be restarted when DFM comes up. If you want it to automatically restart, you must register the adapter as described in the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

What do I have to do if I make a manual change to the Trap Notifier Adapter configuration file?

You must restart the DfmTrapNotifier process for your changes to take effect.

How do I disable the Trap Notifier Adapter?

Select Device Fault Manager > Administration > Fault Notification > Trap Notifier, click Help, and follow the instructions.


Note DFM 1.0 does not offer this function from the GUI. To disable this adapter, you must stop the adapter DfmTrapNotifier process using Server Configuration > Administration > Process Management > Stop Process.


If you have configured the adapter to restart upon CiscoWorks restart and you want to disable automatic restart of the adapter, see the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

HPOV-NetView Adapter

What does the HPOV-NetView Adapter do?
Can the HPOV-NetView Adapter work on a remote HP OpenView or NetView host that is not running CiscoWorks?
If traps are forwarded to HP OpenView or NetView, how are the messages formatted?
What versions of HP OpenView and NetView will work with the HPOV-NetView Adapter?
What do I have to do to get the HPOV-NetView Adapter running?
Under what circumstances would I make a manual change to the HPOV-NetView Adapter configuration file?
Where is the HPOV-NetView Adapter configuration file?
Where is the HPOV-NetView Adapter log file?

What does the HPOV-NetView Adapter do?

The HPOV-NetView Adapter forwards traps from a local or remote HP OpenView or NetView NMS to DFM, enabling DFM to receive traps from devices managed by HP OpenView and NetView.

Can the HPOV-NetView Adapter work on a remote HP OpenView or NetView host that is not running CiscoWorks?

Yes. A special installation script is provided with DFM to do this, as described in the installation guides at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

If traps are forwarded to HP OpenView or NetView, how are the messages formatted?

The format is specified in the DFM MIB, as described in the SNMP Trap Notifier MIB appendix of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

What versions of HP OpenView and NetView will work with the HPOV-NetView Adapter?

The DFM 1.2 and DFM 1.2 Updated for Common Services Version 2.2 HPOV-NetView Adapters work with:

HP OpenView 6.2

NetView:

6.01 (DFM 1.2, and DFM )

7.1 (DFM 1.2 )

The DFM 1.1 HPOV-NetView Adapter works with:

HP OpenView 5.x or 6.x

NetView 5.x (on Solaris and Windows NT)

The DFM 1.0 HPOV-NetView Adapter (previously called the OV/NV Trap Forwarder) works with:

HP OpenView 5.x or 6.x

NetView 6.x

What do I have to do to get the HPOV-NetView Adapter running?

If HP OpenView or NetView is installed on the same machine as DFM, you do not have to do anything. The adapter starts running as soon as it detects HP OpenView or NetView. When CiscoWorks starts, the adapter will automatically start. If HP OpenView or NetView is on a remote machine, you must install the remote HPOV-NetView Adapter on the remote machine, as described in the installation guides at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Under what circumstances would I make a manual change to the HPOV-NetView Adapter configuration file?

You should never change this configuration file.

Where is the HPOV-NetView Adapter configuration file?

The HPOV-NetView Adapter configuration file is at one or both of the following locations, depending on whether you are using HP OpenView or NetView:

HP OpenView: NMSROOT/objects/smarts/conf/OV/server.conf

NetView: NMSROOT/objects/smarts/conf/NV/server.conf

Where is the HPOV-NetView Adapter log file?

The HPOV-NetView Adapter log file is located at one or both of the following locations, depending on whether you are using HP OpenView or NetView:

HP OpenView: NMSROOT/objects/smarts/logs/sm_ov_fwd.log

NetView: NMSROOT/objects/smarts/logs/sm_nv_fwd.log

SNMP Trap Adapter

What does the SNMP Trap Adapter do?
Is the SNMP Trap Adapter the same as the DFM 1.0 Generic Trap Forwarder Adapter?
Where is the SNMP Trap Adapter configuration file?
Where is the SNMP Trap Adapter log file?
What do I have to do to get the SNMP Trap Adapter running?
What if another NMS is already using the trap listening port?
Can I configure the SNMP Trap Adapter to use a different community string for forwarding destinations?
How do I disable the SNMP Trap Adapter?
What do I have to do if I make a manual change to the SNMP Trap Adapter configuration file?

What does the SNMP Trap Adapter do?

The SNMP Trap Adapter listens for SNMP traps sent from devices (or another NMS), and forwards these traps to another network management system (NMS). These are pass-through traps; in other words, DFM does not process these traps.

Is the SNMP Trap Adapter the same as the DFM 1.0 Generic Trap Forwarder Adapter?

Yes; the adapter name was changed after the release of DFM 1.0.

What is the difference between the SNMP Trap Adapter and the Trap Notifier Adapter?

The SNMP Trap Adapter and the Trap Notifier Adapter are similar in that they both forward SNMP traps to other NMSs. The difference is how they treat the traps they receive:

The SNMP Trap Adapter simply treats all received traps as pass-through traps and forwards them.

The Trap Notifier Adapter converts alarms shown on the Monitoring Console into SNMP trap messages (using its own MIB), and then forwards them. (The MIB is described in the SNMP Trap Notifier MIB appendix of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

Where is the SNMP Trap Adapter configuration file?

The SNMP Trap Adapter configuration file is NMSROOT/objects/smarts/conf/trapd/trapd.conf.

Where is the SNMP Trap Adapter log file?

The SNMP Trap Adapter log file is the same as the domain manager log file: NMSROOT/objects/smarts/logs/DFM.log.

What do I have to do to get the SNMP Trap Adapter running?

You normally do not have to do anything to configure the adapter for trap receiving, unless you want to use a listening port other than 162. You do have to configure the adapter for trap forwarding.

For trap receiving:

Verify that the listening port is correct by selecting Device Fault Manager > Administration > Trap Configuration > Trap Receiving, clicking the Help button, and following the instructions.


Note DFM 1.0 does not offer this function from the GUI. To use this adapter, you must edit the configuration file to specify the correct listening port (as described in the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm), and restart the DfmServer process using Server Configuration > Administration > Process Management.


Make sure you have configured your devices (or network management servers) to forward traps to the listening port specified by the adapter (normally 162).

For trap forwarding:

Select Device Fault Manager > Administration > Trap Configuration > Trap Forwarding, click Help, and follow the instructions. You must restart the DfmServer process for your changes to take effect.


Note DFM 1.0 does not offer this function from the GUI. To use this adapter, you must edit the configuration file to specify the correct forwarding port (as described in the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm), and restart the DfmServer process using Server Configuration > Administration > Process Management.


The SNMP Trap Adapter will always restart when CiscoWorks is restarted.

What if another NMS is already using the trap listening port?

If the standard UDP listening port, 162, is already being used, we suggest you use port 9000. You must reconfigure the SNMP Trap Adapter to listen on this port for trap receiving. Also make sure that any NMSs or devices that are forwarding traps use the correct port.

Can I configure the SNMP Trap Adapter to use a different community string for forwarding destinations?

You can do this if you have downloaded and installed DFM 1.2 Patch/IDU 1.2.8 or later (from http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm). For more information, refer to the user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/dfm123cd/u_gd/adapter.htm#1104385.

How do I disable the SNMP Trap Adapter?

You normally do not have to do anything to configure the adapter for trap receiving, unless you want to use a listening port other than 162. For trap forwarding, you do have to configure the adapter.

For trap receiving, you should never disable the adapter for trap receiving.

For trap forwarding:

Select Device Fault Manager > Administration > Trap Configuration > Trap Forwarding, click Help, and follow the instructions. You must restart the DfmServer process for your changes to take effect.


Note DFM 1.0 does not offer this function from the GUI. To disable this adapter, you must edit the configuration file to disable trap forwarding (as described in the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm), and restart the DfmServer process using Server Configuration > Administration > Process Management.


You cannot disable the SNMP Trap Adapter from starting automatically upon CiscoWorks restart, because this would disable trap receiving.

What do I have to do if I make a manual change to the SNMP Trap Adapter configuration file?

You must restart the DfmServer process.

RME Adapter

What does the RME Adapter do?
Where is the RME Adapter configuration file?
Where is the RME Adapter log file?
What do I have to do to get the RME Adapter running?
How do I disable the RME Adapter?
What do I have to do if I make a manual change to the RME Adapter configuration file?

What does the RME Adapter do?

The RME Adapter synchronizes the list of managed devices in a local remote Resource Manager Essentials (Essentials) inventory with the DFM inventory. As soon as a device is added to Essentials, or the configuration of an Essentials device is changed, the device is probed and added to the DFM inventory. However, devices deleted from the Essentials inventory are not automatically deleted from the DFM inventory; you must do this manually. (Unless the device is deleted from the Essentials inventory, it will continue to be re-added to the DFM inventory whenever the inventories are synchronized.)

Where is the RME Adapter configuration file?

There is no such file.

Where is the RME Adapter log file?

The RME Adapter log file is at NMSROOT/conf/dfm/DfmChangeProbe.log.

What do I have to do to get the RME Adapter running?

If Resource Manager Essentials (Essentials) is installed on the same machine as DFM, select Device Fault Manager > Administration > Device Discovery > ChangeProbe, click Help, and follow the instructions. As soon as the adapter detects Essentials, it starts running. You do not need to restart the DfmServer process; upon any CiscoWorks restart, the adapter automatically restarts. If Essentials is on a remote machine, you must install the remote RME Adapter on the remote machine, as described in the installation guides at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

How do I disable the RME Adapter?

Select Device Fault Manager > Administration > Device Discover > ChangeProbe, click Help, and follow the instructions.


Note DFM 1.0 does not offer this function from the GUI. To disable this adapter, you must stop the adapter DfmChangeProbe process using Server Configuration > Administration > Process Management > Stop Process.


If you have configured the adapter to restart upon CiscoWorks restart and you want to disable automatic restart of the adapter, you must unregister the DfmChangeProbe process. Follow the instructions for the notifier adapter processes in the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

What do I have to do if I make a manual change to the RME Adapter configuration file?

There is no such configuration file.

Understanding and Modifying What DFM Console and Window Displays

How do I control what devices and alarms DFM displays in the Alarm Log?
In the Alarm Log, what do the Certainty, Count, Last Change, and First Notify columns represent?
What do the different colors in the Alarm Log mean?
What do the shades of purple (compound) and orange (symptom) in the Alarm Log mean?
Can I change the colors displayed in the DFM consoles?
What does it mean when an attribute is grayed out?
Does DFM limit how many instances of an object it can display?
Time and Timestamp Issues
Clearing and Deleting Alarms

How do I control what devices and alarms DFM displays in the Alarm Log?

To permanently remove all information about an element from the Alarm Log, do the following in the Administration Console:

a. Select the element and right-click Unmanage.

b. Select Inventory > Reconfigure.

c. Select Inventory > Save Inventory.

To display only certain notifications in the Alarm Log, do one of the following:

Use the Alarm Log Filter Criteria dialog box to control everything that is displayed whenever the Alarm Log is opened. From the Alarm Log, select Log >Filter (see the user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

Use a subscription profile to control what is displayed whenever you open the Alarm Log (and other DFM consoles). From the Administration Console, select Even > Maintain Profile (see the user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

In the Alarm Log, what do the Certainty, Count, Last Change, and First Notify columns represent?

Certainty-How reliable the alarm is. Certainty is always 100%, because in DFM fault analysis rule definitions, each alarm is associated with no more than one pattern of events; therefore, certainty is not applicable.

Count-How many times the alarm has occurred since the Administration Console was opened.

Last Change-When the alarm was last generated, relative to the current time.

First Notify-When the alarm was first generated.

What do the different colors in the Alarm Log mean?

Purple indicates a compound event

Orange indicates a symptomatic event

White with blue letters indicates that the event has become inactive (because the event was either addressed or is no longer occurring). After ten polls, inactive events are removed from the display.

White with grey letters indicates that he event state has changed to inactive because DFM no longer subscribes to the element.

What do the shades of purple (compound) and orange (symptom) in the Alarm Log mean?

The darker the shade, the higher the alarm count (in other words, the number of times the alarm has occurred since the console was opened).

Can I change the colors displayed in the DFM consoles?

No; these colors are predefined.

What does it mean when an attribute is grayed out?

A grayed-out attribute indicates one of the following:

The domain manager is not required to poll the device for this attribute value.

The domain manager encountered an error while polling the device for this attribute value. (The device might be down, or it might be up but inaccessible due to network problems.)

The same data is available from another portion of the MIB.

This portion of the MIB is not available on this device model.

Does DFM limit how many instances of an object it can display?

DFM cannot display objects containing more than 1,000 instances. If the number of instances exceeds that limit, you will not be able to open or expand the object when you click its plus (+) sign.

Time and Timestamp Issues

What time zone does a DFM timestamp represent?
Can I change the server timestamp?
What does the information in the general tab of the Notification Properties window represent?
What do the attribute values in the Inventory Browser represent?

What time zone does a DFM timestamp represent?

DFM always uses the server time zone when displaying time information on the consoles. Keep this in mind if your devices use time zones that are different from those of the server.

Can I change the server timestamp?

Yes, by changing the timestamp of the DFM server host machine.

What does the information in the general tab of the Notification Properties window represent?

The following describes what the information in the General tab represents:

The Notification information at the top of the window represents the absolute time of the last generated alarm (the same as Last Notify).

Last Notify represents the absolute time of the last generated alarm
(Current Time - Last Change = Last Notify).

Count represents the number of times the alarm has occurred since the Administration Console was opened.

What do the attribute values in the Inventory Browser represent?

All attribute values in the Inventory Browser represent the current state of the element.

Clearing and Deleting Alarms

How do I remove events I'm not interested in from the Alarm Log?
What happens if I clear an alarm, but other clients (with the same privileges) are viewing the same server's console?
Can I retrieve the value of an attribute that corresponds to when the alarm was generated?
Why does an alarm count continue to increment after I delete the alarm?

How do I remove events I'm not interested in from the Alarm Log?

There are two ways to do this:

For active alarms, select the alarm and right-click Acknowledge. Note that if the alarm continues to be active, it will reappear.

For inactive alarms:

To remove a specific inactive alarm, select the alarm and right-click Delete.

To remove all inactive alarms, select Log > Remove Cleared Events.

What happens if I clear an alarm, but other clients (with the same privileges) are viewing the same server's console?

As long as one console remains open in the session, the alarm will continue to return.

Can I retrieve the value of an attribute that corresponds to when the alarm was generated?

No. The values DFM presents are always current values.

Why does an alarm count continue to increment after I delete the alarm?

The alarm count will increment if the alarm remains active. Remember that the count begins from when the Administration Console was first opened.

Basic Installation Questions

What are the basics I should be aware of when installing DFM?
What must I do to get DFM started?
What happens if I exceed the managed device limit?
How can I find out how many devices/objects DFM is currently managing?
Can DFM coexist with Essentials, Campus Manager, CiscoView, or other Cisco products?
What should I do after installing DFM?

What are the basics I should be aware of when installing DFM?

Use the appropriate version of CD One or CiscoWorks Common Services:

Use DFM 1.2 Updated for Common Services Version 2.2 with either CD One, 5th Edition, or CiscoWorks Common Services 2.2.

Use DFM 1.2 with CD One, 5th Edition.

Use DFM 1.1 with CD One, 4th Edition.

Use DFM 1.0 with CD One, 2nd or 3rd Edition.

If you are installing DFM with the LMS bundle, follow the installation order specified in the appropriate LMS Read Me First document (see http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/cw2000_b/lms/index.htm).

If you want to use DFM with remote versions of Resource Manager Essentials (Essentials), HP OpenView, or NetView, you must install the appropriate adapters on the remote machines. (CiscoWorks need not be installed on the remote HP OpenView or NetView machines.)

If you want your devices to forward traps to DFM, make sure you have configured them to do so.

If another application is using the default listening port for receiving traps, configure the SNMP Trap Adapter to listen on another port.

If you want DFM to forward traps, configure the SNMP Trap Adapter destination port (the recipient machine's server name and port).

Make sure you have configured other applications to receive traps from DFM on the correct port.

Make sure DNS is configured on both the server and client sides.

What must I do to get DFM started?

To get DFM started, you must import some devices.

What happens if I exceed the managed device limit?

DFM will perform slowly and you proceed at your own risk.

How can I find out how many devices/objects DFM is currently managing?

Run the following command:

# NMSROOT/objects/smarts/bin/sm_tpmgr --server=DFM -sizes

Locate the line that is similar to the following:

Number of Ports: 761 [92/92]

In this example, 761 represents the number of discovered ports, out of which 92 are managed. Unless you have reconfigured DFM to manage access ports, you can assume these 92 ports are trunk ports.

Can DFM coexist with Essentials, Campus Manager, CiscoView, or other Cisco products?

DFM can be installed on the same machine as Resource Manager Essentials (Essentials). Check the LMS bundle requirements to verify whether DFM can be installed with any other LMS bundle products (such as Campus Manager and CiscoView). For more information, see the LMS bundle documentation at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/cw2000_b/lms/index.htm. For other Cisco products, check the appropriate product documentation.

What should I do after installing DFM?

Select Server Configuration > Administration > Process Management > Process Status to verify that the following are running:

DfmBroker (the DFM broker process)

DfmServer (the domain manager process)

Adapter processes (depending on which adapters you want enabled):

DfmChangeProbe (the RME Adapter process)

DfmFileNotifier (the File Notifier Adapter process)

DfmTrapNotifier (the Trap Notifier Adapter process)

DfmMailNotifier (the Mail Notifier Adapter process)

Verify that the DFM broker file has been created: NMSROOT/objects/smarts/repos/icf/DFM.rps

Verify that the following log files have been created in NMSROOT/objects/smarts/logs:

DFM.log (the domain manager log)

brstart.log (the DFM broker log)

DfmChangeProbe.log (if you are using the RME Adapter)

Watch the notifications and evaluate whether you should:

Modify any thresholds

Unmanage any objects, or vice versa (for example, you might want to unmanage an interface that is not used and is generating AdministrativelyDown alarms so you do not see these alarms)

Process Management

What are the DFM processes?
What processes automatically start when DFM starts?
Do the DFM processes rely on any other processes?
How do I know if DFM is running properly on a day-to-day basis?

What are the DFM processes?

All CiscoWorks DFM processes are listed in the administration chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

What processes automatically start when DFM starts?

DFM automatically starts the following processes/adapters:

DfmBroker (the DFM broker process)

DfmServer (the domain manager process)

DfmChangeProbe (the RME Adapter process, if Essentials is installed)

HPOV-NetView Adapter (if HP OpenView or NetView is installed)

SNMP Trap Adapter

If you have enabled the notifier adapters, they will not be restarted when CiscoWorks restarts. You can configure them to do so (see the adapter chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm).

Do the DFM processes rely on any other processes?

Yes. When stopping a process, you must first stop any processes that depend on it. These dependencies are listed in the administration chapter of the appropriate user guide at http://www.cisco.com/univercd/cc/td/doc/product/rtrmgmt/cw2000/dfm/index.htm.

How do I know if DFM is running properly on a day-to-day basis?

DFM will not alert you if it goes down. However, if you have installed the Fault History drop-in (available from Cisco.com at http://www.cisco.com/cgi-bin/tablebuild.pl/cw2000-dfm), you will be able to see if the DfmServer process has gone down. (In this case, you might want to change your settings for SNMP timeouts and discovery retries.)