Table of Contents

Best Practices and Sample Configurations

This appendix provides a sample network configuration and offers suggestions to consider when selecting device pairs for monitoring. This appendix will continue to develop to include useful user-based scenarios from which you can model you own site monitoring needs.

This appendix contains:

• Best Practices
  
  
• Useful Cisco IOS Commands
  
  
• Sample Topologies
  
  

Best Practices

Consider the following when setting up SLA device pairs for monitoring:

• Monitor all device connections associated with a primary network connection.
  
  

For example, if an end user at a branch office requires access to a database at the main office for report information to be displayed using HTTP, define the following device pairs to monitor latency:

• Branch office gateway router and main office gateway router
  
  
• Head office gateway router and database server
  
  
• Head office gateway router and Web server
  
  
• Branch office gateway router and Web server
  
  
• Branch office gateway router and database server
  
  

• Reduce network traffic by enabling ME 1100 Series to be installed near the source devices that they are responsible for monitoring.
  
  
• Be aware of administrative overhead costs. Consider installing your ME 1100 Series on your network where Dial on Demand Routing (DDR) is not utilized.
  
  

Useful Cisco IOS Commands

Table C-1 displays Cisco IOS commands to help you with your Cisco IOS device configuration needs. SLM requires SA Agent version 2.1.0 or later, which is supported in Cisco IOS version12.0(5)T or later, and 12.0(8)S or later.

For more information, see the Service Assurance Agent document at http://www.cisco.com/univercd/cc/td/doc/product/software/ios120/120newft/
120t/120t5/saaoper.pdf

Command1	CLI2 Shortcut	Description
Cisco device (config)# rtr responder	-	Enables rtr responder. 3
Cisco device> show rtr application	show rtr appl	Verifies number of operations running. Displays SA Agent version. Requires 2.1.0 or later.
Cisco device> show rtr configuration [entry number]	show rtr conf	Verifies SA Agent is configured. Displays configuration values including all defaults for all SA Agent operations or specified operation.
Cisco device> show rtr responder	-	Verifies if SA Agent enabled on Cisco router. Required to monitor VoIP SA Agent operation.
Cisco device> show rtr collection-statistics [entry number]	show rtr coll	Verifies statistics being collected for SA Agent operation.
Cisco device# debug rtr error [entry number]	-	Enables logging of SA Agent runtime errors.
Cisco device (config)# rtr reset [entry number]	-	Stops all operations and clears SA Agent configuration information. Note   Data will be lost.
Cisco device# service udp-small-servers	-	Enables UDP echo on Cisco routers.
Cisco device> nslookup <ip address> <dns server>	-	Verifies DNS operability.

Note   Entry numbers can be determined by selecting Service Level Management > Administration > Service Level Contracts > View Collection Status from the CiscoWorks2000 Server desktop. Locate the device pair, then click the Status link. The entry number is listed as Data Collector Collection Index.

Sample Command Output

• router# show rtr configuration 3
  
  

        Complete Configuration Table (includes defaults)

Entry Number:3

Owner:slmuser

Tag:AppleTree

Type of Operation to Perform:http

Reaction and History Threshold (milliseconds):5000

Operation Frequency (seconds):60

Operation Timeout (milliseconds):5000

Verify Data:FALSE

Status of Entry (SNMP RowStatus):active

Protocol Type:httpAppl

Target Address:

Source Address:0.0.0.0

Target Port:0

Source Port:0

Request Size (ARR data portion):1

Response Size (ARR data portion):1

Control Packets:enabled

Loose Source Routing:disabled

LSR Path:

Type of Service Parameters:0x0

HTTP Operation:get

HTTP Server Version:1.0

URL:http://www.cisco.com  

Cache Control:enabled

Life (seconds):3600

Next Scheduled Start Time:Start Time already passed

Entry Ageout:never

Connection Loss Reaction Enabled:FALSE

Timeout Reaction Enabled:FALSE

Threshold Reaction Type:never

Threshold Falling (milliseconds):3000

Threshold Count:5

Threshold Count2:5

Reaction Type:none

Number of Statistic Hours kept:2

Number of Statistic Paths kept:1

Number of Statistic Hops kept:1

Number of Statistic Distribution Buckets kept:1

Statistic Distribution Interval (milliseconds):20

Number of History Lives kept:0

Number of History Buckets kept:15

Number of History Samples kept:1

History Filter Type:none

• Complete Configuration Table (includes defaults) show rtr collection-statistics 1
  
  

        Collected Statistics
Entry Number:1
HTTP URL:http://209.165.200.224  
Start Time:*00:01:16.000 UTC Mon Mar 1 1993
 
             Comps:1              RTTMin:343       
             OvrTh:0              RTTMax:343       
        DNSTimeOut:0              RTTSum:343       
        TCPTimeOut:0             RTTSum2:117649  
        TraTimeOut:0              DNSRTT:0         
          DNSError:0           TCPConRTT:13        
         HTTPError:0            TransRTT:330       
          IntError:0            MesgSize:1771      
            Busies:0         

• Entry Number:3 debug rtr error
  
  

May  5 05:00:35.483: control message failure:1
May  5 05:01:35.003: control message failure:1
May  5 05:02:34.527: control message failure:1
May  5 05:03:34.039: control message failure:1
May  5 05:04:33.563: control message failure:1
May  5 05:05:33.099: control message failure:1
May  5 05:06:32.596: control message failure:1
May  5 05:07:32.119: control message failure:1
May  5 05:08:31.643: control message failure:1
May  5 05:09:31.167: control message failure:1
May  5 05:10:30.683: control message failure:1
 

Sample Topologies

• Example I
  
  
• Example II
  
  
• Example III
  
  
• Example IV
  
  

Example I

Example 1 is an enterprise customer that has a head office and 28 district offices (Figure C-1). The backbone includes three Cisco 7200 routers, each running Cisco IOS version 12.0(9)S. There is one DS3 line connecting to the ISP. There is one T1 line connecting to the telco.

The head office provides ISP support to the district offices. A primary objective of this site is to monitor performance between the head office gateway router and each district office.

When considering how to define SLCs and SLA device pairs, the administrator determined:

• Each district office must monitor service to and from the head office gateway router monitor latency.
  
  
• Each district office must monitor service to the web server.
  
  
• Each SLA will contain multiple device pairs that will monitor the SLA in segments.
  
  

TimeSaver Yo can define device pairs between all backbone routers and all 28 district offices (monitoring the same operation type), using the View drop-down list boxes in the Select Devices window. Select your source device type and target device type from the View list, highlight the desired devices, then click Add.The newly defined device pairs are automatically associated so that the first highlighted source device is paired with each highlighted target device, the second highlighted source device is paired with each highlighted target device, and so on.

TimeSaver If you have 28 district offices and a Web server located at the head office, you can define them all as separate device pairs by selecting your source devices using the View drop-down list in the Select Devices window, entering the HTTP server URL in the field provided, then clicking Add.

Note   If you elect to use timesaver instructions, be aware that this could produce a large number of resulting device pairs.

Example II

Example II shows an enterprise organization with a head office and two branch offices (Figure C-2). The head office is equipped with a Cisco 7100 router. Each branch office is equipped with a Cisco 1750 router. A SLM server and an ME 1100 Series are located at the head office. Cisco IOS Release 12.0(5)T is running on each router.

A primary objective of this site is to monitor performance between the head office gateway router and each branch office.

When considering how to define SLCs and SLA device pairs, the administrator determined:

• Each Cisco router must monitor service to the web server.
  
  
• Each branch office must monitor service to and from the head office gateway.
  
  
• Each SLA will contain multiple device pairs that will monitor the SLA in segments.
  
  

The device pairs include:

Note   The ME 1100 Series is a network appliance used to perform distributed data collection and aggregation functions for various network management applications. For more information about the ME 1100 Series, see Installing and Configuring the Management Engine 1100 Series.

Example III

Example III shows an enterprise organization with two branch offices. A Web server is used to transfer files to the branch offices (Figure C-3). Each branch office is equipped with a Cisco 1750 router that provides Internet access. The head office is equipped with a Cisco 7100 that supports remote-office and mobile-office needs, and acts as a gateway. Cisco IOS Release 12.0(5)T is running on each router.

A primary objective of this site is to monitor availability between the Web server and each branch office.

When considering how to define SLCs and SLA device pairs, the administrator determined:

• Each branch office must monitor service to and from the head office gateway.
  
  
• Each district office must monitor service to the web server.
  
  
• Each SLA will contain multiple device pairs that will monitor the SLA in segments.
  
  

The device pairs include:

Example IV

Example IV shows an enterprise organization that has a head office, a regional office, and two branch offices (Figure C-4). The head office is equipped with a Cisco 7100 that acts as a gateway. The regional office is equipped with a Cisco 3600. The branch offices are each equipped with a Cisco 1750 router that provides access to the Internet and which supports voice, data, and fax integration.

A primary objective of this site is to monitor dropped packets between the regional office and each branch office.

When considering how to define SLCs and SLA device pairs, the administrator determined:

• Each branch office must monitor service to and from the regional office gateway.
  
  
• Each SLA will contain multiple device pairs that will monitor the SLA in segments.
  
  

The device pairs include: