|
|
This chapter describes the network management tools you can use with Cisco SES PNNI nodes and PNNI networks, which are as follows:
The following hardware and software components make up the Cisco WAN Manager Network Management workstation:
This section lists the hardware requirements for a Cisco WAN Manager Network Management workstation. Table 11-1 lists the minimum workstation requirements. Using a workstation that meets these requirements will ensure adequate performance.
| Component | Minimum Requirement |
|---|---|
Workstation | Sun Ultra 10 |
Memory | 512 MB |
CPU Speed | 440 MHz |
Hard Disk Drive | 9.1 GB |
Graphics Card | 24-bit |
Monitor | 19 inch |
Three types of machines are supported for WAN Manager 10.2 as standard platforms. They are low-end, mid-range, and high-end platforms. Table 11-2 describes the configuration for each platform.
| Platform Type | Machine Type | Number of CPUs | Size of RAM | Hard Disk Drive | Swap Space | Desktops Supported | Connections Supported |
|---|---|---|---|---|---|---|---|
Low End | Sun Ultra Enterprise 1 Model 151 or Ultra 10 with SCSI controller | 1 | 512 MB | One 9 GB | 1 GB | Less than 5 | Less than 5,000 |
Mid Range option 1 | Sun Ultra Enterprise 2 Model 2300 or Ultra 60 | 2
| 1 GB | Two 9 GB drives | 2 GB | 5 - 10 | 5,000 - 50,000 |
Mid Range option 2 | Sun Ultra Enterprise 2 Model 2300 or Ultra 60 | 2
| 2 GB
| Two 9 GB drives | 2 GB | 10-20 | 50,000 - 100,000 |
High End | Sun Enterprise 4000 or Enterprise 450 | at least 4 | 4 GB | Two 9 GB drives or disk array | 4 GB | More than 20 | More than 100,000 |
 |
Note The minimum CPU speed requirement for all but the low end platforms is 300 MHz. All platforms require a 24-bit graphics card. |
The selection of a proper CWM platform depends on a number of factors, such as, the number of CWM desktops, the number of managed connections, and the number of statistics collected and stored. Table 11-2 lists recommended CWM platforms based on the size of network.
The following are additional notes for CWM platform requirements:
This section lists the required software to install on the Cisco WAN Manager Network Management workstation to manage the SES PNNI nodes and PNNI networks.
Refer to the appropriate chapters in the Cisco WAN Manager Installation Guide for Solaris, Cisco WAN Manager 10.2 for general workstation setup (including disk partitioning) and installation procedures.
A change in the installation procedure requires that you use the following disk partitioning requirements instead of those found in the CWM 10.2 Installation Guide for Solaris (Doc-7810308=).
Sufficient disk space and proper disk partitioning are essential to achieving the best performance from CWM and your network management workstation.
|
Note The minimum disk space requirement for CWM 10.2 is one 9-GB disk drive. |
Use the following commands to gather some of the required information. The dmesg command provides information about the workstation type, amount of memory and CPU speed.
dmesg | moreThe format command enables you to determine information about the disk drives on your workstation. Select a disk from the list of those available, and use the verify command to determine the current partitioning of each disk.
This section describes how to partition a CWM workstation's single 9-GB disk drive. The following procedure ensures that all but the final partition will be at a set size.
|
Note The actual total disk space for a 9-GB disk varies depending on the manufacturer of the disk. One 9-1 GB disk might have 9.05 GB of available space, while another might only have 8.9 GB of space available. Other disk drive flaws also limit disk capacity. Most disks will NOT have a full 9.1 GB, and the slice 2 (s2) total will vary. |
| Slice | Partition | Space | Comments |
|---|---|---|---|
s0 | / | 2000 MB | Allocate third |
s1 | swap | 1030 MB | Allows for memory upgrade to 1 GB; allocate second |
s2 | <overlap> | 8996 MB | Total amount of space on the disk; do not attempt to modify |
s3 | /opt | 500 MB | allocate fourth |
s4 | /var | 1000 MB | allocate fifth |
s5 |
|
|
|
s6 | /usr/users | 2500 MB | Must be 2000 MB; allocate first |
s7 |
| 1966 | Raw partition; might be less that this amount; allocate last |
|
Note To check the running total of what disk space is remaining, just click on any other partition to update the total free value. When the total free is 0 free and a rounding error of 0 or 1, click OK. |
This section describes how to partition a CWM workstation that has two 9-GB disk drives. When you install Solaris, partition the first disk drive as shown in Table 11-4. Do not partition the second disk. The second disk is automatically partitioned during the CWM software installation.
| Slice | Partition | Space | Comments |
|---|---|---|---|
s0 | / | 2000 MB | Allocate third |
s1 | swap | 1030 MB | Allows for memory upgrade to 1024 MB; allocate second |
s2 | <overlap> | 8996 MB | Total amount of space on the disk; do not attempt to modify |
s3 | /opt | 1000 MB | Allocate fourth |
s4 | /var | 1000 MB | Remainder of disk; allocate last |
s5 |
|
|
|
s6 | /usr/users | 2000 MB | Must be 2000 MB; allocate first |
s7 |
|
|
|
|
Note The total disk space should equal the space shown in s2. |
The following technical documents comprise the CWM 10.2 documentation set:
For SES PNNI networks using in-band management, provide the following information for your network:
NETWORK:Network2 GATEWAYS:sj234567 DISCOVERY PROTOCOL:PNNI
|
Note To save your changes while using the vi editor, remember to press Esc, colon (:), then wq!. |
Use the Topology Configurator to provide information required to communicate with the nodes. Also use the configurator to specify Network IP.
|
Note These are nodes that have their SNMP community string for GET operations not set to public, and SNMP community string for SET operations not set to private. |
To configure a PXM-SES card, telnet to the card and execute the following commands:
pxmses1> shmsimulateresetReason 0 pxmses1> deltree "C:DB2"
|
Note The above commands should be used when inconsistency exists between the database and the image. |
pxmses2> addcontroller 2 i 2 7 "pxmses controller
|
Note The step above is only required after a dbtree or clrallcnf command. |
pxmses2> addpnport 9:1.1:1 assumes a BXM in slot 9 pxmses2> cnfpnportsig 9:1.1:1 -nniver pnni 10
Cisco WAN Manager (CWM) provides the following features for the Cisco SES PNNI controller:
Table 11-5 lists the types of SPVCs supported by the SES PNNI controller using CWM 10.2.
| Endpoint 1 | Endpoint 2 | SPVC Connection Type | Node Type |
|---|---|---|---|
BXM | BXM | ATM SPVC | Routing nodes on BPX with feeder node on SES |
The Cisco SES PNNI controller is managed through WAN CiscoView 3.2. WAN CiscoView 3.2 requires Release 5.1 of the CiscoView Engine, which is included on the CWM 10.2 CD.
CiscoView Release 5.1 has been migrated from X/MOTIF to a JAVA based application. WAN CiscoView 3.2 supports line, port and resource partition configuration and real-time counters on Cisco SES PNNI controllers.
The look and feel of CiscoView 5.1 is slightly different from CiscoView 4.2, but most dialog screens will be familiar to experienced CiscoView users. Rear view selection (normally done by selecting the outer part of the device with the right mouse button) is not available for the Cisco SES PNNI controller.
During the CWM 10.2 installation process, CiscoView 5.1 and BPX-SES device packages are installed for you. This eliminates the need to incrementally select device packages to install.
Accessing CiscoView is a simple task. From the CWM 10.2 Topology Map, a device can be selected for management by CiscoView.
When you start CiscoView, the CiscoView main window opens. The following components comprise the CiscoView main window:
Use the Select Device drop-down list box to select and display a device. Either enter a device name or IP address, or select from the recently displayed devices listed.
Device names and SNMP read and write community strings are preserved when you open new CiscoView sessions.
Use the Device Commands buttons to activate device commands unique to the displayed device.
The Device Command buttons are described in the online help for each device package.
Use the Main Menu buttons to perform various CiscoView tasks.
Use the Graphical Device Display window to view a graphical display of the device's back or front panel once you select a device. The display shows all device components color-coded according to their current status and refreshed according to your polling frequency. If a hot swap is detected, the device is rediscovered and the display redrawn at the next poll.
Use the Status Bar and buttons to display the result of device polling, selections, and so on.
Launches a Telnet command-line session to the managed device.
Launches a separate browser containing the Cisco Connection Online (CCO) web page.
This feature is not supported in WAN CiscoView 3.2.
Opens the TAC Mailer dialog box for sending reports to the Cisco Technical Assistance Center (TAC) group. You can describe the problem using the available options and the comment field. When you click Send, your descriptions and information about the runtime device package and operating environment are sent to the specified mail recipients.
This feature is not supported in WAN CiscoView 3.2.
Opens the Preferences dialog box where you can specify SNMP and community string. The preferences settings are preserved for all new CiscoView 5.1 sessions.
Displays the following:
CiscoView release version and copyrights
Active device package, if applicable
All installed device package information
Opens CiscoView 5.1 help if no device is selected.
Opens context help if a device or component is selected.
This feature is not supported in WAN CiscoView 3.2.
Displays the progress and result of device polling, selections, and so on.
Displays system information (name, description, location, contact, and up-time) for a displayed device.
Prints the current graphical display.
Describes the significance of the colors on the graphical display. Color schemes are:
When you select a device in CiscoView, a graphical representation of the device is displayed. You can view front device panel and select different components and menu options to configure and monitor status for these devices.
Configures device categories, such as Node Management, NNI, and so on.
Displays a set of dynamic charts for selected device categories.
Displays either the front or back device panel. The BPX-SES has only a front panel view.
Reduces the graphical display down to 90%, 80%, 70%, 60%, or 50%. You can resize it back up to 100%.
Triggers component polling and display update.
Displays system MIB information (name, description, location, contact, and up-time) for a device.
Once you have installed CiscoView and learned to navigate within it, you can perform various tasks.
Depending on your platform, you can start CiscoView:
Select a device to view its graphical representation to configure and monitor it. The device names and SNMP read and write community strings are preserved when you open new CiscoView sessions.
Use the Set Preferences option to change certain options within CiscoView.
Select a component on the graphical device display to configure and monitor it.
Use the Configure menu to configure multiple categories of information, for example, Interface, Management, Physical, and ARP Table, simultaneously.
Different categories of information can be displayed for each device, card, and port. To see the categories of information that can be displayed for each component type, look at the Category pop up menu from the Configuration window.
Use the Monitor menu to monitor multiple categories of information, for example, Ethernet collisions, Management, Physical, and ARP Table, simultaneously. The Monitoring dialog is non-modal and resizeable.
Use the Preferences Community tab to delete the read and write community strings for the device currently being managed. This lets you enter the read and write community strings for a device after you display the device. If you want to make changes to a device or port setting, but did not specify community string when you first opened the device display, you can enter the community string without exiting and reopening the device window.
If a host's community strings are not already defined within CiscoView, you can add them with the CiscoView Community Strings dialog. Otherwise, CiscoView allows you to enter the correct community strings when you try to access the host.
If you do not enter a host's community strings when accessing the host, CiscoView uses the default read and write community strings of "public" and "private".
Use the Preferences SNMP tab to set polling frequency, SNMP timeout and retries, and default read and write community strings. The recommended values for preferences are:
Use the Default Read and Write Community fields to define the community strings that CiscoView automatically uses for device when you do not specify the device's current community strings.
Writes modification of all categories to managed device then closes the dialog box
Writes modification of the current category to managed device, leaving the dialog box open
Aborts changes and closes the catalog list
Prints the current category
Launches device-specific help
Launches a table row creation dialog box
Deletes a selected row from the table
Use the Device Support Utility to integrate new Cisco device information asynchronously with the CiscoView engine, uninstall device packages, install new device packages, or upgrade existing installed packages.
The Device Support Utility operates in one of two modes: Interactive mode or Command Line mode. The functionality of both modes is similar; the only difference between the two is that Interactive mode provides a Graphical User Interface (GUI). Each mode allows the user to display a list of currently installed device packages and their versions, uninstall one or more packages, and automate device package installations and upgrades.
Use the Device Support Utility to:
From a UNIX platform, you can start the Device Support Utility by running the script "xdsu" from the ~svplus/wancv/bin directory.
In Interactive mode, the Install Device Packages dialog box installs new device packages or upgrades existing packages. The Device Support Utility will not allow you to select a package whose superseding version has already been installed in the package repository.
In Interactive mode, the Device Support Utility dialog box shows a list of the device packages that are already installed. It also acts as a launch point for uninstalling device packages.
The following information describes how to test the basic connectivity and setup for CiscoView. Perform these tasks first when you have a CiscoView-related problem.
From the UNIX workstation, try to ping the router's IP address. If the ping is unsuccessful, make sure that IP routing is properly enabled and is functioning. Use "ping -s" to check for slow IP response. Ping the device by its Network IP as well as by its LAN IP address.
If you can ping the device by its LAN IP address but not its Network IP address, there is a Network IP problem. Consult your system administrator for assistance in resolving this problem.
Enter the dspsnmp command to view the SNMP configuration and verify the community strings. If the strings are not correct, configure the device with the cnfsnmp command.
Use the Preferences SNMP tab to set polling frequency, SNMP timeout and retries, and default read and write community strings. The recommended values for preferences are:
Any Cisco SES PNNI controller configuration which is provisioned via provisioning commands can be shown using the SES CLI show commands. The SES CLI show commands are described in detail in Appendix B, "SVC, SPVC, and PNNI Commands".
You can find port information with the following show commands:
You can find out SSCOP information with the following show commands:
You can show information about specific calls with the following Show commands:
The Cisco SES PNNI controller supports the Call Trace feature as two distinct facilities, as defined by ATM Forum PNNI v2.0 Living List (July `98):
Both these facilities can be used to trace a call in the Control Plane.
The Connection Trace facility can be used to determine the path taken by an existing connection---Point-to-Point, or Point-to-Multipoint---from any node in the network to the destination node of the call. You initiate the trace of a connection through the SES CLI by providing the ingress interface, the call reference for a p2p call, and in addition endpoint reference for p2mp call. This generates a TRACE CONNECTION message, which includes Trace Transit List (TTL) IE. This message would always travel towards the node which has the called address.
Each node fills up TTL IE with node ID and Egress logical port ID and passes the message on. The egress logical port Id is obtained from call record of the existing connection, using CallRef and EndPtRef (p2mp). The destination node makes the portId zero.
This feature is not supported on an IISP interface, neither for the transporting of IEs nor for the processing of the IEs.
The destination node copies the TTL IE as is into TRACE CONNECTION ACK message, and sends it back to the source node with status - `trace completed normally'. You can then view the Trace Transit List to find out the path taken by an existing connection.
A connection trace may fail at any node for the following reasons:
1. Message is dropped as it is not supported.
2. Feature is disabled and the node refuses to participate.
3. Trace can not be progressed to the destination node due to some failure (say the call is already released, or call release is in progress).
You are then informed of these failure events, with appropriate cause values along with the TTL up to the last node which sent the failure message back.
See "SVC, SPVC, and PNNI Commands" for command syntax details.
The syntax of the connection trace (conntrace) command is given in Appendix B, "SVC and PNNI Commands."
Figure 11-1 shows the section within which a call can be traced. For instance, you could start the trace for a call setup between SRC and DST. The Trace Connection message would terminate at the IISP interface and an ACK/NACK would be sent back (similar to as if its a UNI). This would enable the user to complete a partial trace.
Path Trace facility allows you to trace calls in real time. You enable or disable this feature node wide, on a per UNI interface basis or based on called party/calling party number. When enabled, the source node adds a TTL IE as part of the Setup/AddParty message and subsequent nodes supporting this feature add their own TTL IE. Various flags can be turned ON/OFF as part of the TTL on the source node, enabling a user to filter details on the trace. These flags are:
1. Hierarchy (H): Information from all the DTLs in the hierarchy are added if this flag is enabled, as defined in the ATM Forum specifications.
2. Crankback (CB): If a call fails with a Crankback, the cause value is inserted in the TTL IE.
3. Call Clear/Retain (X): The call is cleared immediately after reaching the destination node if the flag is enabled by sending Release/AddParty Reject and so on, else it is established by sending Connect. These messages contain the TTL IE with proper status codes.
|
Note This feature is not a safeguard against mis-configuration, to prevent the calls from being released inadvertently. |
4. VPI/VCI values (V): If enabled, VPI/VCI values of the egress port are filled up in the TTL IE at every node.
5. Call Reference values (CR): If enabled, Call Reference values of all the egress ports are filled up in the TTL IE.
Path trace information for all the traced calls are stored in the trace log file, where they can be display information on the CLI screen.
The information from the response message is displayed using show commands. This information includes the TTL IEs and result of the return code.
Path trace fails due to similar reasons specified above for Connection Trace Fail. However, since this is a real-time trace, the node detecting the failure fills in the proper information in TTL IE, apart from filling up the Release Message.
An IISP interface is treated similarly to a UNI interface and the trace is terminated.
The Pathtrace commands are as follows:
Their syntax is described in Appendix B, "SVC, SPVC, and PNNI Commands."
If path trace is enabled, the source node adds the TTL IE in the Setups depending on the flags set. The via node processes this IE and adds the relevant octets into the IE.
The signaling stack checks this flag before decoding TTL IE in an incoming setup message. It then checks this flag before starting TTL IE encoding procedures for an outgoing call.
This command is for incoming calls on a source node. Signalling checks these flags/parameters before generating/decoding the TTL IE. If the user selects disable, the rest of the parameters are ignored.
This command is used for interoperability, when the Cisco SES PNNI controller is connected to equipment which does not support path trace. It takes two options, described below.
Figure 11-2 illustrates a sample network with pathtrace insert and remove IEs.
The Cisco SES PNNI controller supports Connection Trace and Path trace facility for maximum 10 hops network and maximum 5 traces are allowed to be in progress at a time. The trace log file is recycled every time it reaches 1MByte in size.
Posted: Thu Jul 27 19:49:13 PDT 2000
Copyright 1989-2000©Cisco Systems Inc.