Installation guide

Network Infrastructure for EtherNet/IP™
Deploying the Network
5-62
The IP address of the switch can generally be configured using a console cable directly connected to the
device. Alternatively, a Dynamic Host Configuration Protocol (DHCP) or Bootstrap (BootP) server can
be used.
Managed switches also allow for remote management of the network. There are multiple network
management software vendors that provide solutions to manage Faults (alarms) remotely, Configuration,
Accounting (who is using what resources), Performance, and Security—commonly referred to as
FCAPS.
Problems identified during infrastructure testing will be detailed by the network tester so appropriate
corrective action can be taken. Other non-conformance problems and warnings may appear that address
additional issues, such as incorrect termination, faulty components, and high ambient noise. All non-
conformance and test warning issues must be resolved before network startup.
5.2 Putting Traffic on the Network
In order to identify features of the EtherNet/IP network infrastructure that help provide the required
performance and connectivity, it is necessary to characterize network traffic, employ IP addressing
techniques, and understand IP multicast traffic.
5.2.1 Characterization of EtherNet/IP Traffic
As has been shown, EtherNet/IP network infrastructure can be defined as a hierarchical interconnection of
Layer 2 and Layer 3 Ethernet switches.
Traffic generated during programming, configuration, and diagnostics of EtherNet/IP devices as well as
during exchange of non-time-critical data is called
explicit messaging in EtherNet/IP terminology. It is
normally low-rate traffic that generally has an insignificant impact on network performance. Although it
contains both broadcast and unicast data, this traffic does not require engagement of any special features
in the EtherNet/IP network infrastructure. Broadcast traffic typically consists of IP packets supporting
Address Resolution Protocol (ARP), BootP, DHCP, SNMP and other protocols of this type. Unicast
traffic consists of Transmission Control Protocol/Internet Protocol (TCP/IP) packets.
Traffic generated during time-critical data exchange is called
implicit messaging and consists mostly of
User Datagram Protocol/Internet Protocol (UDP/IP) unicast and multicast packets. Examples include:
• Input/Output (I/O) data and status produced by a remote I/O device for consumption by
one or more programmable controllers
• Data produced by a programmable controller for consumption by one or more
programmable controllers
While the handling of UDP/IP unicast traffic does not require engagement of any special features in the
EtherNet/IP network infrastructure, handling of UDP/IP or IP multicast traffic may require these features.
Although EtherNet/IP supports change-of-state and application reporting, in a typical control system, data
exchange is predominately cyclic. The implicit messaging traffic is normally generated at an aggregate
rate of tens of thousands of packets per second, depending on the number and type of EtherNet/IP devices
and the application. Many EtherNet/IP devices are, for example, capable of generating up to 5000 packets
per second. Normally, this traffic is evenly divided between UDP/IP unicast and multicast packets. Packet
length is typically around 120 bytes.
Normal IP multicast traffic generated in an EtherNet/IP network consists of high-rate, short-packets
generated on a continuous basis. For this reason, EtherNet/IP networks differ considerably from typical