Routing is one of the fundamental parts of an infrastructure that keeps the network operational. It is therefore important that the routing protocols are sufficiently secured to minimize tendencies of compromising it.
Super-W is a large retail company based in Mid West U.S. The company has traditionally operated with a single corporate campus location and a single distribution center serving 20 retail locations. Currently, Super-W has a single corporate campus and 4 large distribution centers that support more than 200 retail outlets. In this respect, internetwork is necessary for an upgrade to support the expanded remote locations with security at the core of the process.
In my capacity as the network engineer in charge of Super-W upgrade, I will design an upgrade procedure that details the processes, protocols, routing topology and justification for a 99.99% uptime network. My time comprises of five network analysts and a technical project manager. This paper describes the best upgrading procedure that factors in seamless migration with minimal disruption to the customer and business operations. It will also recommend a routing protocol adopted in this scenario that will deliver the best in terms of performance and security. Further, the routing topology will be illustrated graphically for better comprehension. Finally, availability to a degree of 99.99% is expected. Class four availability of the network implies that the system should be 99.99% up and running every time. Unavailability of 0.01% or 1.01 minutes per year will be favorable for the business.
UPGRDING PROCEDURE
Super-W is proud to present the complete upgrade of its core network infrastructure to support more business and remote locations. This is in tandem with the recent acquisition of the four large distribution centers that support more than 200 retail outlets. The upgrade will commence with a resource limit of $500,000 and a team of four network analyst and one technical project manager. The deployment of the infrastructure based on Cisco Switching Systems will result in significant enhancements for the employees and the customers. This may include;
- Increased network redundancy and reduced time for network failover
- Increased network capacity
- Support for numerous Gigabit ports
- Support for full native IPv6
- Support for ever expanding routing tables
Over the course of the upgrade all the Super-W customer access switches, transport and transit circuits will be migrated to the new core infrastructure. The process will proceed by first moving customer access switches starting Monday 10.00PM through 2.00AM until Friday. The team has spent considerable resources and time devising a mechanism that minimizes customer impacts. However, considerable impact will still need to be felt to ensure success of the system. Customers are advised that they will experience a series of 3-10 minutes outages when the particular customer access switch that maintains their equipment is transferred to the new core network infrastructure. Once the transfer is complete other additional maintenance will be scheduled and announced as appropriate to move the transit and transport circuit to the new infrastructure. Thus, in summary, the upgrade will commence in three phases, namely;
- Customer access switch upgrade
- Transit circuit upgrade
- Transport circuit upgrade
With regards to IPv6 service, once the upgrade of all the circuits has been completed to the new network, the IPv6 configurations will be accorded as required by each remote entity. It is expected that the upgrade procedure will be completed in the first half of April 2014.
ROUTING PROTOCOL
As the network engineer, I feel confident recommending Enhanced Interior Gateway Routing Protocol for Super-W. This is because in my view, EIGRP can be implemented much quicker and with a fraction of the cost required for other upgrades such as OSPF. EIGRP will also provide the functionality that will accord the stabilization that Super-W desires. Implementing EIGRP provides a scalable platform and stable environment and Super-W can desire to implement OSPF in future if need be.
EIGRP
The Enhanced Interior Gateway Routing Protocol is a routing protocol developed by Cisco Systems and Introduced with Cisco Internetworking Operating System (Cisco IOS). It combines the advantages of link-state protocols such as OSPF and distance vector protocols such as IGRP. Enhanced IGRP utilizes Diffusing Update Algorithm for quick convergence.
OSPF is an Interior Gateway Protocol developed for use by Internet Protocol based internetworks. OSPF distributes routing information between routers belonging to an autonomous system which subsequently exchange routing information via a common routing protocol. OSPF is based on shortest path-first as well as link-state technology.
Super-W contains of a single campus network and 200 distribution centers. An efficient routing protocol is necessary to ensure operational efficiency as well as security. In the present scenario, EIGRP is chosen over OSPF. This is because given the budget of $500,000, EIGRP is less costly and easy to implement than OSPF. Likewise, since earlier protocols utilized by the company were based on Cisco, it is important to keep up with it to achieve economic significance. EIGRP qualifies as the best choice in this scenario because of its link-state and distance vector protocol elements.
EIGRP
This routing protocol supports IP, Novel NetWare, and AppleTalk. Since it is a nonhierarchical topology, it summarizes subnet routes of directly connected networks at a network number boundary. Route summarization is possible where subnet routes of directly connected networks are automatically summarized at network number boundaries.
Enhanced IGRP employs bit-wise sub-netting and variable-length sub-network masks (VLSM’s) to save address spaces and achieve companies addressing in a corporate internetwork.
Route selection in IGRP is possible through the selection of route metrics to compare the best routes from a selection of possible routes. Enhanced IGRP separate metric values assigned fro bandwidth, delay, reliability, and load. By default, Enhanced IGRP computes the metric for a route using the minimum bandwidth of each hop in the path and includes a media specific delay in the hop. Generally, the metrics used by Enhanced IGRP includes the following;
- Bandwidth
- Delay
- Reliability
- Load
EIGRP can best be implemented in this case using as a mesh network topology comprising of 200 routers of which are connected to 2 to 4 neighboring routers.
The Enhanced IGRP uses a DUAL convergence algorithm. This ensures that convergence is achieved quickly and efficiently. It uses two mechanisms;
Fig. 1 Sample DUAL convergence
First, Enhanced IGRP router keeps its neighbors routing table to allow it to use a new route to a destination instantly if a feasible route is known. If a feasible route is not found based on the previous routing information, the router running the EIGRP automatically becomes active for that destination and sends out queries to its neighbors requesting for alternate routes. These queries will propagate until a final destination is found. Routers unaffected by the topology change need not be involved in the query and response. Second, a router using Enhanced IGRP receives full routing tables from its neighbors during the first contact with them. Thereafter, only changes to the routing tables are executed by routers directly affected by the change. Successors (neighboring routers) are used for packet forwarding providing the least cost route to the destination.
In the figure above, DUAL convergence is calculated based on destination N. Each node shows its cost to destination N while the arrows indicate the nodes successor Thus, C uses A to reach N and the cost is 2. If the link between A and B fails, B initiates a query informing A, C, D, and N that it has lost a feasible successor. Upon receiving the request, D determines if it has any pother feasible successor and if negative, it starts a route computation and enter the active state. C is feasible successor because it costs less than, for instance D to destination N. D can switch to C as its ultimate successor. C and A are unaffected in this case because they did not participate in the change.
In case of a route computation, we consider C and A. C acknowledges that it has lost a feasible successor and has no other successor. Thus, C has to perform route computational and therefore sends a query to its neighbors (actually its only neighbor D). D is prompted to reply because its successor is still intact and when C receives the reply, it acknowledges that all neighbors have processed the feedback in relation to failure to N. At this point D is chosen by C as the new feasible successor to reach N with a cost of 4. A and B were unaffected in this case.
Enhanced IGRP find feasible successors by computing the following network scalability
- Router memory usage
- CPU
- Bandwidth
- Security
Router security is essential for availability of services and overall network performance. Enhanced IGRP security is a feature provided in Cisco routers to prevent accidental or malicious routing disruptions caused by hosts in the network. Additional controls such as route filters are used to prevent inappropriate learning or propagating routing information.
Thus, the adoption of Enhanced IGRP routing protocol for Super-W is intended to provide the following advantages over other types of protocols;
Improved router memory and CPU utilization
Intelligent bandwidth control possible through bandwidth calculation prior to transmitting updates
Greater flexibility for wireless connections where timing intervals must be fine-tuned to a particular device or bandwidth
EIGRP requires less cost in terms of manpower and timed. Since only five personnel and $500,000 are available for the whole upgrade process, this option will ensure that the process is completed in due time with exact cost estimates.
In achieving fourth class availability, downtime should be significantly reduced while uptime should be maintained. Availability is expressed in percentage of reliability. For instance a 99.99% reliable Super-W network will only be down 1.01 minutes a week or 4.32 minutes a month. This percentage of reliability is expressed in terms of uptime, downtime, and response and recovery time. For uptime to be upheld, response and recovery time should be significantly reduced, subsequently increasing uptime. The routing infrastructure suggested above has all the features that ensure that the system is highly redundant to attacks that may result in downtime and increased response and recovery time. .
References
Doyle, J. (2012). Routing TCP/IP Volume I (CCIE Professional Development). Cisco Press.
http://www.cisco.com/cgi-bin/Support/PSP/psp_view.pl?p=Internetworking:EIGRP .
Pepelnjak, I. (2013). EIGRP Network Design Solutions: The Definitive Resource for EIGRP Design, Deployment, and Operation. Cisco Press.
Retana, A. (2012). EIGRP for IP: Basic Operation and Configuration. Addison Wesley.
