One problem with broadcasting is that lots of broadcast traffic on a LAN can slow network traffic
down, as well as slow individual systems down. If there is so much broadcast traffic on the LAN
that other non-broadcast traffic is significantly delayed (or never delivered), this is called a
broadcast storm. Broadcast storms typically arise when network loops are created through faulty
network configuration, but can also happen as the result of a malicious attack. For example, a
classic Denial of Service attack is to send an ICMP echo request ("ping") over the LAN that
specifies the source address of a system and a broadcast address for the destination. Every
system receiving the ping will respond to it -- flooding the system specified as the source of the
ping with ICMP echo replies.
There are also other security concerns associated with broadcasting. Since all the systems in the
broadcast domain can see broadcast packets, the information in them is susceptible to discovery,
intercept, and modification. This is of particular concern in industrial Ethernet environments
(where, for example, manufacturing processes are controlled directly by computers) and in any
environment (such as government and finance) where sensitive data is regularly transmitted over
the LAN.
A number of methods can be used to mitigate problems and threats associated with large
broadcast domains, including broadcast filtering and physically separating large broadcast
domains into smaller domains. The problem with these solutions is that the are typically
implemented at the Network Layer (Layer 3), and require Layer 3 devices (such as routers and
firewalls) to implement them. These Layer 3 devices require separate subnets, and themselves
emit a significant amount of broadcast traffic.
What we really want is a way of abstracting the idea of a LAN so that large broadcast domains can
be separated into smaller domains
without requiring any network rewiring or physical movement
of systems
. We’d also like the ability to extend broadcast domains across Layer 3 devices to
physically remote systems.
With a VLAN, the broadcast domain for a particular system is determined by the software settings
on the Layer 2 switch port to which the system is connected.
So, for example, in a traditional LAN, all the systems connected to Switch A would be part of
Broadcast Domain A. If the switch is a VLAN-capable switch, then it is possible to configure
several ports on the switch for VLAN A, several others to VLAN B, others to VLAN C, and so on.
This allows you to both:
l
reduce the number of devices in local broadcast domains
l
extend broadcast domains across devices separated by more than one switch
The predominant VLAN standard is 802.1q. This standard adds a VLAN tag to the information in the
Ethernet packet. Since they operate at the switching level, VLANs are Layer 2 technologies --
though they are often confused with Layer 3 subnetting, because in many configurations there is
one VLAN configured per subnet. This is usually done for the practical purpose of allowing the
systems on a VLAN to be managed as a group by other network management devices/software
that work by IP address ranges, for example, rather than VLAN tags.
Copyright © 2014 Coyote Point Systems, A Subsidiary of Fortinet, Inc.
All Rights Reserved.
111
Equalizer Administration Guide
Summary of Contents for Equalizer GX Series
Page 18: ......
Page 32: ...Overview 32 Copyright 2014 Coyote Point Systems A Subsidiary of Fortinet Inc ...
Page 42: ......
Page 52: ......
Page 64: ......
Page 72: ......
Page 76: ......
Page 228: ......
Page 238: ......
Page 476: ......
Page 492: ......
Page 530: ......
Page 614: ......
Page 626: ......
Page 638: ......
Page 678: ......
Page 732: ...Using SNMP Traps 732 Copyright 2014 Coyote Point Systems A Subsidiary of Fortinet Inc ...
Page 754: ......
Page 790: ......
Page 804: ......
Page 842: ......
Page 866: ......