JUNOS High Availability

Chapter 12. Managing Intradomain Routing Table Growth

This chapter provides a useful collection of methods, tips, and tricks for controlling routing table growth. While controlling the size of routing tables was once imperative because of memory and processor limitations on many vendors’ routing platforms, those limitations are no longer the prime reasons to do so. Because modern networks’ routing tables are minimized as much as possible to simplify operation and troubleshooting, controlling routing table growth is a much more basic equation—fewer routes equals fewer potential routing issues, and therefore increases uptime.

For example, suppose you came into the operations center to bring up a new local area network (LAN) with a bank of web servers during a routine 2:00 a.m. maintenance window. Shortly after bringing the servers online, you lose remote connectivity to the LAN. The flashing red icon on the network management map confirms the loss and you immediately reverse the changes. You then spend a few hours trying to figure out why the activation of the subnet impacted traffic. After much digging, you determine the issue to be the result of several small factors, the classic “death by a thousand cuts,” but it’s already 5:00 a.m. The maintenance window closed an hour ago. So, what caused the failure?

• A few static routes that were no longer needed were left configured on a router.

• An IP address conflict existed between the new server LAN and several existing LANs.

• The crisis was ...