The Continuing Evolution of OSPF

Previous Table of Contents Next


The solution would be to allow the OSPF interface found in R1 and R2 that are the area 0 links to also belong to NSSA 1. Under the current specification, each routers’ interface can only belong to one area. This draft creates a new parameter known as secondary areas with the original area becoming the primary area. Using Opaque LSAs, routing information is provided between participants.

This Internet Draft is highly recommended for additional reading as it has a lot of merit. Hopefully this Draft will be advanced to an RFC very soon.

OSPF Optimized Multipath (OSPF-OMP)

Date Published: March 1998

Author: Curtis Villamizar

Expiration Date: September 1998

File Name: draft-ietf-ospf-omp-00.txt

Abstract

OSPF may form multiple equal cost paths between points. This is true of any link-state protocol. In the absence of any explicit support to take advantage of this, a path may be chosen arbitrarily. Techniques have been utilized to divide traffic somewhat evenly among the available paths. These techniques have been referred to as Equal Cost Multipath (ECMP). An unequal division of traffic among the available paths is generally preferable. Routers generally have no knowledge of traffic loading on distant links and, therefore, have no basis to optimize the allocation of traffic.

Optimized Multipath is a compatible extension to OSPF, utilizing the Opaque LSA to distribute loading information, proposing a means to adjust forwarding, and providing an algorithm to make the adjustments gradually enough to ensure stability yet provide reasonably fast adjustment when needed.

Draft Summary

The author of this draft does an excellent good of explaining the need for this draft, and he goes through the various type of ECMP. He covers the three different techniques used in ECMP, as shown in the following list:

  Per packet round-robin forwarding.
  Dividing destination prefixes among available next hops in the forwarding entries.
  Dividing traffic according to a hash function applied to the source and destination pair.

Additional discussion then ensues on how and why they fall short in properly providing the required load-sensitive routing to demonstrate the need for the Internet Draft.

To expect any protocol to perform load-sensitive routing, it is necessary for the protocol to gain information on the different possible paths to the required destination. This Internet Draft proposes the use of Opaque LSAs that will flood the load information. After this information is received by OSPF, it can calculate the correct route. The Opaque LSA will be responsible for providing the correct route calculation. In addition, Opaques LSAs will provide the following information:

  Measure of link loading in each direction as a fraction of link capacity
  Measure of packets dropped due to queue overflow in each direction (if known) expressed as a fraction
  Link capacity in kilobits per second (or unity if less than 1,000 bits per second)

The information provided in the following list comes from sampling interface counter values that are available via SNMP:

  Bytes Out
  Bytes In
  Packets Out
  Packets In
  Output Queue Drops
  Input Queue Drops

Chapter Summary

This chapter discussed IETF’s Working Drafts, specifically those that apply to OSPF and its future. This information will enable you to be proactive in understanding how the upcoming changes to OSPF are going to affect your organization’s network. It is important to note yet again that these drafts are considered works in progress and as such they can change at any time.

Case Study: NetFlow Switching

NetFlow Switching is part of Cisco’s new Internet of Quality Services initiative. It enables enterprise networks to meet and exceed many critical benefits by providing the following enhancements:

  Flexible and detailed accounting, billing, and charge back for network and application resource utilization, including dedicated line- and dial-access accounting.
  NetFlow provides both proactive network monitoring and trouble resolution. It allows for successful monitoring and profiling of customer network usage for use in the future.
  Responsive tactical network engineering and strategic network planning. The information collected by this tool can be used by NetSys.
  Excellent outbound marketing and customer service via intimate knowledge of customer network and application usage patterns. It allows detailed information based upon applications used by the customer to be examined in great detail.

What is NetFlow exactly? NetFlow is defined as a sequence of packets in one direction between given source and destination endpoints, which are identified by IP address. These flows can be extremely granular.

The difference between regular network switching and NetFlow Switching is that regular switching handles incoming packets with separate serial tasks for switching, services, and traffic measurements. The typical process involved with switching a packet is shown in Figure 11-2.


Figure 11-2  Regular network switching.

With NetFlow Switching, however, the process is applied to only the first packet of a flow. Information is extracted from the first packet and is used to build an entry in the NetFlow cache for this flow. Subsequent packets are handled via a single streamlined task, which handles switching, services, and data collection concurrently. Figure 11-3 illustrates the processes of NetFlow Switching.


Figure 11-3  NetFlow Switching.

NetFlow Requirements and Deployment Strategy

NetFlow does not require the adoption of any new or proprietary protocols or new networking equipment. NetFlow may be deployed in an extremely flexible manner to include interface-by-interface deployment.

NetFlow does require a server upon which a software program referred to as Flow Collector is run. This software collects the data and stores it for use by network engineers. It has a variety of features to optimize the data collected based upon user requirements. This software also acts as a management platform of sorts for routers with the NetFlow capability. These Flow Collectors are designed to be deployed throughout the network and all will report to a central station referred to as the Central Flow Analyzer. It is this flow analyzer that enables the network engineers to manipulate the data and reports from the data collected by the flow collectors.


Previous Table of Contents Next




OSPF Network Design Solutions
OSPF Network Design Solutions
ISBN: 1578700469
EAN: 2147483647
Year: 1998
Pages: 200
Authors: Tom Thomas

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net