08.OSPF Planning and Design Process

OSPF Planning and Design Process Cisco CCNP ROUTE Training Instructor: Joe Rinehart, CCIE #14256

OSPF Planning and Design Process Cisco CCNP ROUTE Training

In This Lesson: Unique Characteristics of OSPF When OSPF is the Best Design Choice  OSPF Planning Methodology  Verification Plan for Implementation  Implementation Steps Transitioning to OSPF from Another Protocol  Transitioning  


Unique Characteristics of OSPF 1. Brief Brief Histor History y of of OSPF OSPF 2. Link-Sta Link-State te Protocol Protocol Chara Characteri cteristics stics 3. Unique Unique OSPF OSPF Feat Feature uress


Brief History of OSPF Need for More Robust IGP IETF Formed OSPF Working Group in 1988 to Replace RIP: Network size limitations Inability to choose higher b andwidth interfaces/links Derived from Several Sources Bolt, Beranek, and Newman's (BBN) SPF algorithm for ARPANET (1978) Radia Perlman’s research on fault tolerant broadcast of routing data (1988) BBN’s Area Routing Work (1986) Early Version of OSI IS-IS Protocol

1987

OSPF Working Group Formed

•

1989

First OSPF Interoperability Testing 1991 OSPF Demo at Interop ‘91 1993

Internet Adopts CIDR 1995

OSPF V1 Published

–

–

OSPF V2 Published OSPF Recommended IGP

•

–

OSPF V2 Updated Crypto Added

–

1997

OSPF V2 Updated

–

John Moy, OSPF: Anatomy of an InternetRouting Protocol (Reading Mass: Addison-Wesley, 1998)

–


Link-State Protocol Characteristics Network 1

Formal Neighbor Relationships Hello packets sent for neighbor discovery and maintenance Database synchronization and exchange Active tracking of neighbor state Link State (Topology) Database Constructed of information about active paths (links) in the network All routers contain copies of the database Update Process Once database is built only changes sent •

Network 2

•

R1

R2

•

•

R3

R4

Link Outage

X

R3 LSA

•

•

Network 3

•

Network 4

•

When there is a change it is flooded Dijkstra algorithm used for calculation


Link-State Protocol Characteristics Network 1

Rapid Convergence Link-state database and path calculation process is rapid and very efficient Loop avoidance inherent in the protocol Higher CPU and memory utilization Complexity Greater amount of effort required in planning and design process Configuration requires more work •

Network 2

•

R1

R2

•

•

R3

R4

Link Outage

X

Network 3

•

•

R3 LSA

Network 4

Vastly greater depth in routing logic and computation process


Unique OSPF Features Network 1

Formal Hierarchy Ability to segment the network into multiple section called areas Area 0 (Backbone) is required and all other areas must directly connect to it Flooding and route calculation is restricted within each area Route Summarization Support for aggregation/supernetting Configurable on a per-area basis Industry Standard Protocol “Open” designates OSPF as standard Implemented by every equipment vendor •

Network 2

•

R1

R2

•

R3

R4

Link Outage

X

•

•

R3 LSA

Network 3

•

Network 4

•


Unique OSPF Features Network 1

Cost as Basis for Route Selection Bandwidth of interface assigned a cost Lowest cumulative cost to destination network makes for best route Different Network Types Supported Broadcast (e.g., Ethernet LAN) Non-Broadcast Point-to-Point Point to Multipoint (Non)Broadcast Neighbor Types Designated Router (DR): Single peer on broadcast network Backup DR: Secondary peer •

Network 2

•

R1

R2

•

•

•

•

R3

R4

Link Outage

X

R3 LSA

•

•

Network 3

Network 4

•


When OSPF is the Best Design Choice 1. In Multi-Vendor Routing Environment 2. For Scalability in Large Networks 3. Design Principles for OSPF


When OSPF is the Best Design Choice In Multi-Vendor Routing Environment Routing and/or Layer 3 Switches Can Be from ANY Equipment Manufacturer OSPF V2 (IPV4) and V3 (IPV6) assumed Open standard •

–

–

•

Vendor Support List (Partial) Cisco Avaya Juniper HP Adtran Enterasys –

–

–

–

–

–


When OSPF is the Best Design Choice For Scalability in Large Networks OSPF can experience stuck-in-active conditions in very large topologies Link State Database synchronization and flooding behavior limited to areas Address summarization in areas Fewer SPF calculations using areas •

•

•

•


When OSPF is the Best Design Choice Design Principles for OSPF Multi-Access Network Peering Routers establish one peering relationship (DR/BDR) Select and configure DR/BDR using the priority command Serial Interface Bandwidth OSPF also assumes 1.544 Mbps for any serial interface Use the bandwidth command to set the actual value Remember that frame-relay is inherently a non-broadcast medium •

–

–

•

–

–

–


When OSPF is the Best Design Choice Design Principles for OSPF Limit Number of Routers per Area Older documentation recommends 50 Depends on a variety of factors: Type of area (stub, NSSA, etc.) Number of external LSA’s How well other areas are/can be summarized CPU and memory available on the devices Media type (LAN/Frame-Relay, etc.) •

–

–

•

•

•

•

•


When OSPF is the Best Design Choice Design Principles for OSPF Limit Number of Area Border Routers ABR’s keep separate databases for each area connected Also depends on a variety of factors Can be avoided by allocating connectivity to additional devices Goal is not to overload the ABR Summarize Addressing When Possible Design and use IP addressing that can be summarized •

–

–

–

–

•

–

–

Deploy address summarization whenever practical and possible


When OSPF is the Best Design Choice Design Principles for OSPF Carefully Choose Area Designs Avoid placing devices in area 0 unless critical to overall network operations Use redundant links between backbone routers whenever possible Maintain consistency in designation of labeling for areas Create a template for areas Avoid the use of virtual-links Try to anticipate growth in the design process •

–

–

–

–

–

–


OSPF Planning Methodology 1. 2. 3. 4.

General Network Design Principles High Level Design Pilot/Proof of Concept Testing Review and Verification Process


OSPF Planning Methodology General Network Design Principles Start with REQUIREMENTS Know what the expectations are Be able to clearly explain the expected outcomes Gather all of the facts Perform a physical site inspection Account for power, cabling, and device placement Know the availability and timeline of WAN circuits •

–

–

–

–

–

–

Account for any existing devices that require support

–


OSPF Planning Methodology General Network Design Principles Document all ASSUMPTIONS Financial (capital budget/operational costs) Who is responsible for what Expected time for design, implementation and testing Other people/organizations involved Ask questions for confirmation •

–

–

–

–

–

–

–

–

Get everything in writing and signed as confirmation Develop contingency plans Develop and stick to a project plan


OSPF Planning Methodology High Level Design Create High-Level Diagram(s) Site Locations Proposed equipment VLAN/LAN IP Addressing Wide Area Connection Links Routing protocol domains Develop a Bill of Materials (BOM) Equipment (type and quantity) Lead time/delivery dates SMARTnet Maintenance Staging logistics •

–

–

–

–

–

•

–

–

–

–


OSPF Planning Methodology High Level Design System-Wide IP Addressing Allocation Subnet, mask(s), equipment IP’s Consistency in addressing Database of addresses and locations Reserved and unused ranges documented Ability to summarize contiguous ranges of addresses Track usage/decommissioning of addresses within the pool(s) •

–

–

–

–

–

–

•

Turn the high-level plan into detailed plans after approvals


OSPF Planning Methodology Pilot/Proof of Concept Testing When Implementing New Technologies, a Pilot or Proof of Concept Trial is Advisable Technology: Wireless, Unified Communications, etc. Devices: New models/types of LAN switches, routers, etc. Acquisition: Business merger, telecommuter program, etc. Protocol: Change of routing protocol, etc. Process: Changes the way the company is doing business (e.g., replacing card payment systems) •

–

–

–

–

–


OSPF Planning Methodology Pilot/Proof of Concept Testing Map Out a Testing Environment that Represents the Larger Project Scope: Subset of the number and type(s) of sites Devices: Same type(s) of devices to be used in larger implementation Staging: Create a lab-centric miniature version of the network Testing Plan: Create a plan that will test the limits and outcomes Reporting: Codify results and conclusions for use in the larger implementation (“lessons learned”) •

–

–

–

–

–


OSPF Planning Methodology Review and Verification Process Involve as many competent people in the process of reviewing your designs and documentation •

–

–

–

–

–

Peers: Those in your department Experts: Individuals whose skills and experience are greater than your own Critics: Carefully chosen people whose evaluation is likely to be accurate Superiors: Direct supervisor or department heads Others: Additional reviewers as needed


Implementation Steps 1. Detailed Project Plan 2. Device-Specific Configurations 3. Verification Steps


Implementation Steps Detailed Project Plan Basic Document Format Identification of stakeholders Identification of roles and responsibilities Timelines and milestones Communication plan Account for power, cabling, and device placement Implementation plan(s) Expected outcomes Criteria for completion •

–

–

–

–

–

–

–

–


Implementation Steps Device-Specific Configurations Routers/Layer 3 Devices Hostnames (adhering to standards) Interface configurations Security configurations EIGRP Routing configuration elements Usernames/Passwords/TACACS Network Time Protocol SNMP/Syslog WAN configurations •

–

–

–

–

–

–

–

–

Device access Advanced configurations

–

–


Implementation Steps Device-Specific Configurations Layer 2/3 Switches Hostnames (adhering to standards) SVI Interface configurations Security configurations Usernames/Passwords/TACACS Network Time Protocol SNMP/Syslog •

–

–

–

–

–

–

VLAN configurations Device access Advanced configurations

–

–

–


Verification Steps Commands Verifying Configuration(s) General Connectivity/Functionality show ip interface brief show cdp neighbors show environment all show diag show inventory OSPF Routing show ip protocols show ip ospf interfaces show ip ospf neighbor show ip ospf database show ip route ospf •

–

–

–

–

–

•

–

–

–

–

–


Verification Steps Commands Verifying Configuration(s) General Connectivity/Functionality debug ip packet debug interface OSPF Routing debug ip ospf adjacency debug ip ospf events debug ip ospf hello debug ip ospf spf debug ip ospf lsa-generation •

–

–

•

–

–

–

–

–

–

–

debug ip ospf packet debug ip ospf flood


Transitioning to OSPF from Another Protocol 1. In a New (“Greenfield”) Environment 2. In an Existing Network


Transitioning to OSPF from Another Protocol In a New (“Greenfield”) Environment Assumptions Not necessarily a first-time network installation May involve new equipment and/or new telecommunications circuits •

–

–

–

–

Probably maintaining identical LAN addressing Other routing protocols in use in all or part of the existing network Static routing EIGRP RIP •

•

•


Transitioning to OSPF from Another Protocol In a New (“Greenfield”) Environment Transition Process Stage new equipment with OSPF Connect only a small test network for analysis purposes Have full configuration available for day of transition Have full testing plan codified Schedule outage window after hours Migrate area 0 First Disconnect old devices and swap with new OSPF devices Execute test plan and certify network •

–

–

–

–

–

–

–

–


Transitioning to OSPF from Another Protocol In an Existing Network Assumptions Existing equipment will host OSPF No new circuits or IP addressing Involves a “hot cut” to implement Requires several key elements: Transition plan Testing plan •

–

–

–

–

•

•

•

Roll-back plan in the event of issues


Transitioning to OSPF from Another Protocol In a New (“Greenfield”) Environment Transition Process Configure OSPF with a higher administrative distance (suggested distance of 224) Ability to establish OSPF neighbor relationships and peering May stage “dummy” IP addresses to serve as testing points Schedule an outage window after hours At transition time set administrative distance back to 110 Execute test plan and certify network •

–

–

–

–

–

–


Key Terms You Should Know Link-State — Class of Interior Gateway Protocol that relies on all devices having an identical understanding of the network (links) DR/BDR — Designated Router/Backup Designated Router, the primary/secondary devices that all devices on a multi-access establish a neighbor relationship with Area — In OSPF, a partitioned section of the network in which all routers maintain an identical copy of the link state database Area 0 — Also referred to as the backbone area, the OSPF area required area through which all traffic must transit Roll-Back Plan — Contingency plan for restoring the network to its original state following an unsuccessful transition attempt Greenfield — Term used to describe installation of a brand-new network infrastructure in a given location


What We Covered  Unique Characteristics of OSPF   When OSPF is the Best Design Choice   OSPF Planning Methodology   Verification Plan for Implementation   Implementation Steps   Transitioning to OSPF from Another Protocol 

08.OSPF Planning and Design Process

Recommend Documents