Configuration Management: Best Practices White Paper: Document ID: 15111
Configuration Management: Best Practices White Paper: Document ID: 15111
Paper
Document ID: 15111
Introduction
High−Level Process Flow for Configuration Management
Create Standards
Software Version Control and Management
IP Addressing Standards and Management
Naming Conventions and DNS/DHCP Assignments
Standard Configuration and Descriptors
Configuration Upgrade Procedures
Solution Templates
Maintain Documentation
Current Device, Link, and End−User Inventory
Configuration Version Control System
TACACS Configuration Log
Network Topology Documentation
Validate and Audit Standards
Configuration Integrity Checks
Device, Protocol, and Media Audits
Standards and Documentation Review
Related Information
Introduction
Configuration management is a collection of processes and tools that promote network consistency, track
network change, and provide up to date network documentation and visibility. By building and maintaining
configuration management best−practices, you can expect several benefits such as improved network
availability and lower costs. These include:
We have seen the following issues resulting from a lack of configuration management:
This best−practice document provides a process flowchart for implementing a successful configuration
management plan. We'll look at the following steps in detail: create standards, maintain documentation, and
validate and audit standards.
Create Standards
Creating standards for network consistency helps reduce network complexity, the amount of unplanned
downtime, and exposure to network impacting events. We recommend the following standards for optimal
network consistency:
Note: Similar network devices are defined as standard network devices with a common chassis providing a
common service.
• Determine device classifications based on chassis, stability, and new feature requirements.
• Target individual software versions for similar devices.
• Test, validate, and pilot chosen software versions.
• Document successful versions as standard for similar−device classification.
• Consistently deploy or upgrade all similar devices to standard software version.
The first step to successful IP address management is understanding the IP address blocks used in the
network. In many cases, network organizations have to rely on RFC 1918 address space, which isn't Internet
addressable, but can be used to access the network in conjunction with Network Address Translation (NAT).
Once you have defined the address blocks, allocate them to areas of the network in a way that promotes
summarization. In many cases, you'll have to further subdivide these blocks based on the number and size of
subnets within the defined range. You should define standard subnet sizes for standard applications, such as
building subnet sizes, WAN link subnet sizes, loopback subnet size, or WAN site subnet size. You can then
allocate subnets for new applications out of a subnet block within a larger summary block.
For example, let's take a large enterprise network with an east coast campus, a west coast campus, a domestic
WAN, a European WAN, and other major international sites. The organization allocates contiguous IP
classless interdomain routing (CIDR) blocks to each of these areas to promote IP summarization. The
organization then defines the subnet sizes within those blocks and allocates sub−sections of each block to a
particular IP subnet size. Each major block or the entire IP address space can be documented in a spreadsheet
showing allocated, used, and available subnets for each available subnet size within the block.
The next step is to create standards for IP address assignments within each subnet range. Routers and Hot
Standby Router Protocol (HSRP) virtual addresses within a subnet might be assigned the first available
addresses within the range. Switches and gateways may be assigned the next available addresses, followed by
other fixed address assignments, and finally dynamic addresses for DHCP. For example, all user subnets may
be /24 subnets with 253 available address assignments. The routers may be assigned the .1 and .2 addresses,
and the HSRP address assigned the .3 address, switches .5 through .9, and the DHCP range from .10 through
.253. Whatever standards you develop, they should be documented and referenced on all network engineering
plan documents to help ensure consistent deployment.
• Creates a consistent access point to routers for all network management information related to a
device.
• Reduces the opportunity for duplicate IP addresses.
• Creates simple identification of a device showing location, device type, and purpose.
• Improves inventory management by providing a simpler method to identify network devices.
Most network devices have one to two interfaces for managing the device. These may be an in−band or
out−of−band Ethernet interface and a console interface. You should build naming conventions for these
interfaces related to the device type, location, and interface type. On routers, we strongly recommend using
the loopback interface as the primary management interface because it can be accessed from different
interfaces. You should also configure loopback interfaces as the source IP address for traps, SNMP and syslog
messages. Individual interfaces can then have a naming convention that identifies the device, location,
purpose, and interface.
We also recommend identifying DHCP ranges and adding them to the DNS, including the location of the
users. This may be a portion of the IP address or a physical location. An example might be
"dhcp−bldg−c21−10" to "dhcp−bldg−c21−253", which identifies IP addresses in building C, second floor,
wiring closet 1. You can also use the precise subnet for identification. Once a naming convention has been
created for devices and DHCP, you'll need tools to track and manage entries, such as Cisco Network
Registrar.
We recommend creating standard configurations for each device classification, such as router, LAN switch,
WAN switch, or ATM switch. Each standard configuration should contain the global, media, and protocol
configuration commands necessary to maintain network consistency. Media configuration includes ATM,
Frame Relay, or Fast Ethernet configuration. Protocol configuration includes standard IP routing protocol
configuration parameters, common Quality of Service (QoS) configurations, common access lists, and other
required protocol configurations. Global configuration commands apply to all like devices and include
parameters such as service commands, IP commands, TACACS commands, vty configuration, banners,
SNMP configuration, and Network Time Protocol (NTP) configuration.
Descriptors are developed by creating a standard format that applies to each interface. The descriptor includes
the purpose and location of the interface, other devices or locations connected to the interface, and circuit
identifiers. Descriptors help your support organization better understand the scope of problems related to an
interface and allows faster resolution of problems.
We recommend keeping standard configuration parameters in a standard configuration file and downloading
the file to each new device prior to protocol and interface configuration. In addition, you should document the
standard configuration file, including an explanation of each global configuration parameter and why it is
important. Cisco Resource Manager Essentials (RME) can be used to manage standard configuration files,
protocol configuration, and descriptors.
Upgrade procedures may vary widely depending on network types, device types, or new software
requirements. Individual router or switch upgrade requirements may be developed and tested within an
architecture group and referenced in any change documentation. Other upgrades, involving entire networks,
can not be tested as easily. These upgrades may require more in−depth planning, vendor involvement, and
additional steps to ensure success.
You should create or update upgrade procedures in conjunction with any new software deployment or
identified standard release. The procedures should define all steps for the upgrade, reference vendor
documentation related to updating the device, and provide testing procedures for validating the device after
the upgrade. Once upgrade procedures are defined and validated, the upgrade procedure should be referenced
in all change documentation appropriate to the particular upgrade.
Solution Templates
You can use solution templates to define standard modular network solutions. A network module may be a
wiring closet, a WAN field office, or an access concentrator. In each case you need to define, test and
document the solution to help ensure that similar deployments can be carried out in exactly the same way.
This ensures that future changes occur at a much lower risk level to the organization since behavior of the
solution is well defined.
Create solution templates for all higher−risk deployments and solutions that will be deployed more than once.
The solution template contains all standard hardware, software, configuration, cabling, and installation
requirements for the network solution. Specific details of the solution template are shown as follows:
• Hardware and hardware modules including memory, flash, power, and card layouts.
• Logical topology including port assignments, connectivity, speed, and media type.
• Software versions including module or firmware versions.
• All non−standard, non device−specific configuration including routing protocols, media
configurations, VLAN configuration, access lists, security, switching paths, spanning tree parameters,
and others.
• Out−of−band management requirements.
• Cable requirements.
• Installation requirements including environmentals, power, and rack locations.
Note that the solution template does not contain many requirements. Specific requirements such as IP
addressing for the specific solution, naming, DNS assignments, DHCP assignments, PVC assignments,
interface descriptors, and others should be covered by overall configuration management practices. More
general requirements, such as standard configurations, change management plans, documentation update
procedures, or network management update procedures, should be covered by general configuration
management practices.
Maintain Documentation
We recommend documenting the network and changes that have occurred in the network in near real−time.
You can use this precise network information for troubleshooting, network management tool device lists,
We recommend creating a cross−functional team to measure configuration management success and improve
configuration management processes. The first objective of the team is to implement configuration
management performance indicators in order to identify configuration management issues. We'll discuss the
following configuration management performance indicators in detail:
After evaluating the results from these audits, initiate a project to fix inconsistencies and then determine the
initial cause of the problem. Potential causes include a lack of standards documentation or a lack of a
consistent process. You can improve standards documentation, implement training, or improve processes to
prevent further configuration inconsistency.
We recommend monthly audits, or possibly quarterly if only validation is needed. Review past audits to
confirm that past problems are resolved. Look for overall improvements and goals to demonstrate progress
and value. Create metrics to show the quantity of high−risk, medium−risk, and low−risk network
configuration inconsistencies.
Cisco RME is a configuration management tool that can audit and report on hardware versions, modules and
software versions. Cisco is also developing more comprehensive media and protocol audits that will report
inconsistency with IP, DLSW, Frame Relay and ATM. If a protocol or media audit is not developed, you can
use manual audits, such as reviewing devices, versions and configurations for all like devices in a network, or
by spot checking devices, versions and configurations.
You should review the following documentation on a quarterly basis: standard configuration definitions,
solution templates including recommended hardware configurations, current standard software versions,
upgrade procedures for all devices and software versions, topology documentation, current templates, and IP
address management.
Related Information
• Technical Support − Cisco Systems
All contents are Copyright © 2006−2007 Cisco Systems, Inc. All rights reserved. Important Notices and Privacy Statement.