NETWORK DESIGN IN

PROFINET
OPTIMAL PERFORMANCE
FROM THE FIRST TELEGRAM

WHITE PAPER
CONTENTS

ABOUT INDU-SOL
Indu-Sol is a technology company
with international operations. It
is a key supplier of reliable net-
work infrastructure in production
(operational technology/OT) and
industrial IT (IIT).

The company also has many years

of experience in diagnostics for
data communication quality. This
means that Indu-Sol can offer
customers conceptual support in
the network planning stage, and
provide, as a partner, practical
knowledge and solutions through-
out the entire network life-cycle.

Indu-Sol 2 White paper

1
DO TOO MANY COOKS
REALLY SPOIL THE BROTH?
How planners, designers, and operators work
Page 4

2
together to achieve a stable PROFINET network

SEARCH AND YOU WILL FIND! Page 6

Selected causes of unstable network communication

3
THE SWITCH AS CENTRE-
PIECE OF NETWORKED
STRUCTURES
Page 12

4
CENTRAL CONDITION MONI-
TORING BY SOFTWARE
Page 16

5
Bringing together distributed diagnostic data

CONCLUDING THOUGHTS Page 18

ON A SUSTAINABLE NETWORK
STRUCTURE
Indu-Sol 3 White paper
1
DO TOO MANY COOKS
REALLY SPOIL THE BROTH?
How planners, designers, and operators work
together to achieve a stable PROFINET network

Nowadays, machines and systems in industrial produc- What factors can actually endanger stable operation?
tion are not isolated units. Rather, they are connected to The “Field Service Excellence” working group from the user
one other by networks of varying complexity. As different organisation PROFIBUS & PROFINET International (PI),
as the perspectives and priorities of planners, designers and Indu-Sol GmbH measurement engineers, undertake
and operators, etc., involved in this system may some- a great number of service calls annually to evaluate the
times be, they are all united by a common objective: The communication quality in industrial networks. In the ma-
machines and systems – often designed for a life-span of jority of cases, it is a troubleshooting. The communication
10 to 15 years or longer – should produce continuously quality had already noticeably worsened and the affected
from day one; and the products should satisfy the high- company could not solve the problem itself, and so had
est quality standards. The “too many cooks” proverb can to call in external support. When the service technician
be proved wrong in this case if all the participants work arrived, there was often no diagnostic data available. This
together to establish the prerequisites for stable network lengthened the service call unnecessarily. The VORTEX re-
operation. In addition to the right network structure, this port from Indu-Sol GmbH summarises the annual findings
especially involves selecting infrastructure components on the possible causes of deteriorations in communication
based on needs. quality in industrial networks:

Other

Excessive cable length EMC

Software & devices

Cables

Connectors

Source: VORTEX 2019, Indu-Sol GmbH, Page 9.

Connec- Ex. cab. Software &
Cause Cables EMC Other Data basis: 454 service calls made by Indu-Sol in 2018. The
tors length devices
main cause of the quality reduction as identified during the
Proportion
30% 28% 20% 10% 4% 8% service call is always stated. Download the document at:
in 2018
www.indu-sol.com/pn19-vortex

Indu-Sol 4 White paper

 This white paper has been developed using this know­ Perfect PROFINET communication is the
ledge and Indu-Sol GmbH's 17 years of experience in the basic requirement for stable machine
objective evaluation of communication quality in indus- and/or system operation. In order to
trial networks. We will present important quality param- identify quality deteriorations before
eters for stable data communication, and corresponding they become visible, the causes (e.g.
monitoring options that must already be considered in the worn plugs/cables, electromagnetic
planning phase, provided in the design, and used when couplings, etc.) and their effects have
operations are running. Today, PROFINET is one of the to be monitored.
most frequently used automation technologies. Its use has
been growing steadily for several years. For this reason,
the white paper will focus on this technology.

PROFINET ‒
THE MOST IMPORTANT FACTS
■■ Ethernet-based automation technology
■■ Data transmission up to max. 100 Mbps
■■ Simultaneous sending and transmission of data possible (Full-duplex mode)
■■ Organisation of data communication via switches, which, as intelligent data intersections,
transmit the information to suitable receivers
■■ Cyclic data volume controllable by setting the update rate per device
■■ Explanatory video: This is how PROFINET works. Watch online now:
www.indu-sol.com/pn19-film

Indu-Sol 5 White paper

2
SEARCH AND YOU WILL
FIND!
Selected causes of unstable network
communication

The car is a good example of how the intelligence of

technical devices can help to quickly detect anomalies:
When cars were just beginning there was no pre-warning
of the failure of individual components due to wear. Fill
levels of oil, cooling water, tyre pressure, etc. – had to
be regularly checked by hand. Later on there were at the components in industrial networks are also subject
least yellow and red lamps on the dashboard. Nowadays, to wear from usage. If you add to this harsh production
standard diagnostics are displayed as written notifica- environments (heat/cold, pressure, chemicals, vibra-
tions, e.g., “Fall in pressure! Check tyre pressure”, “Fill tions, mechanical strain), then cables, connectors and
windscreen wiper water”, or “Service required”. The re- so on, will at some stage reach their limits. But when
sult is that car owners only react when they need to and is this point reached? And is it better to then replace
know immediately what action to take. Just like cars, everything just in case?

Quality parameters at a
glance: Which values influence
the stability of PROFINET
communication?
In PROFINET, telegrams are checked by the switch at verification are thus already the basis for stable data
the latest during transmission. Faulty telegrams – caused communication.
by abraded contacts or cable breaks – are discarded and The VORTEX report mentioned previously, and statistics
counted as errors in the port statistics of the switch. from the user organization PROFIBUS & PROFINET Inter-
This means that they never reach their destination. Even national (PI/“Field Service Excellence” working group),
if error telegrams do not lead directly to a machine/ also underline the huge importance of electromagnet-
plant standstill, they are a clear indicator of what will ically compatible machine and plant installation for
happen if measures are not taken. The quality of the stable operation. Among other things, the rise in the
machine/plant installation and its measurement-based use of higher frequency power electronics (e.g. frequency

Indu-Sol 6 White paper

inverters) is increasing the current load in industrial net- In spring 2018, the user organisation
works. Since the shields of the data lines usually have a PROFIBUS & PROFINET International
lower impedance than the equipotential bonding system, (PI) published a new guideline on
they provide a more advantageous path. functional-equipotential bonding and
The resulting vagrant currents therefore use the shields shielding in PROFIBUS and PROFINET
of the data lines as a return current path instead of the networks.
intended green-yellow PE conductor. In this way, they are
diverted via the devices/switches. An up-to-date equi-
potential bonding design thus protects the data lines
along with the connected devices. Whether or not this
objective is achieved can be proven by measurement.
This could involve leakage current monitoring integrated
in the switch, for example.
protective and functional-equipotential bonding in the
area of machine and plant construction: And this not
without having to discard, as already mentioned, many
old ways of doing things.
In contrast, the guidelines on values with regard to
network limit in PROFINET networks have existed since
2014. The data transmission speed in PROFIBUS was still
limited to 12 Mbps, but with PROFINET it was increased
to 100 Mbps, and for good reason: The increasing net-
working of machines and plants, an increasing level of
automation, and the hunger for data in maintenance or
quality control, for example, mean a growing data vol-
This has been known for many years. Nonetheless,
for a long time there were no set rules when it came ume and consequently, a higher network load. As an Eth-

to grounding and shielding in the OT area, as any writ- ernet-based technology, PROFINET also enables different

ten specifications that did exist were technically very protocols to communicate via the same infrastructure.

outdated. With the PI guideline “Functional Bonding As well as cyclic process data traffic required for con-

and Shielding of PROFIBUS and PROFINET”, the field- trolling the automation processes, there is also acyclic
bus user organization has for the first time attempted data traffic (e.g. BDE data, touch panel commands, or
to present on paper the current state-of-the-art re- TCP/IP data of a monitoring camera), which all increase
garding protective and functional grounding, as well as the network load.

Indu-Sol 7 White paper

2
SEARCH AND YOU WILL
FIND!
Selected causes of unstable network
communication

In order to ensure a permanently stable network The network limit thus has a huge influence on the
operation in spite of this, the user organisation performance of the network. In many cases, it also ex-
­PROFIBUS & PROFINET International (PI) specifies in plains “mysterious” events (sporadic events that cannot
the ­
PROFINET D
­esign and Commissioning Guideline be reproduced) in the network whose occurrence seems
that the basic load in the network should not be al- to lack a definite pattern. As a result of acyclic data
lowed to exceed 20%: traffic, temporary peak loads can occur that prevent

Network time-critical data from being transmitted on time. Acy-

Recommendation
limit clic data traffic causes inconsistent network utilisation
< 20% No action required – temporary peak loads can lead to the loss of current

Review of planned network limit is data.

20% ... 50%
recommended

> 50% Take action to reduce network limit Tip:

Source: PROFINET planning guidelines from PI, page 118 Monitoring the network limit permanently
during operation makes sense as you can
It therefore makes sense to plan the cyclic network uncover sporadically occurring peak loads.
limit at the design stage, and to monitor it permanently Pay attention to the monitoring interval
during subsequent plant operation. though: Peak loads sometimes only occur
Why is acyclic traffic (red line of carriages in image) in the millisecond range! A network limit
so critical? Some telegrams, for instance, TCP/IP data evaluation in the second range necessar-
of a monitoring camera, are very large and in this way ily works with average values. Peak loads
block the transmission paths for time-critical telegrams occurring in milliseconds are averaged out
(green cars in the image), which then do not arrive on and therefore remain hidden.
time.

Source: Scene from “PROFINET − The film | Technology in simple terms”

Indu-Sol 8 White paper

How can the possible causes
of events be identified
securely?
It is common when diagnosing Port statistics
PROFINET networks to use the LED
Overview Netload m/s Discards Error

on the controller as an indicator

for the system status. If the LED Port 1 Port 2 Port 3
9% 24% 16%
Partner: 10.1.9.2 Partner: 10.1.9.29 Partner: 10.1.9.15
lights up red then it is certainly
time to act. The controller, how-
Port 4 Port 5 Port 6
1% 35% 2%
ever, only performs device diag- Partner: 10.1.9.23 Partner: 10.1.9.16 Partner: 10.1.9.22

nostics, and not communication

Port 7 Port 8 Port 9
or network diagnostics. If all the 3% 0% 0%
Partner: 10.1.9.24 Partner: - Partner: -

devices are intact but the com-

munication between them is not
optimal or the infrastructure components are affected, The web interface of PROmesh switches − here the PROmesh P9 − is clearly laid
out and reveals correlations between diagnostic results: The network loads at
this is not identified by the red LED. In addition, the ports 2 and 5 (image above) are too high. This means that telegrams have to be

controller in PROFINET does not perform any preventative discarded (Discards, image below).

diagnostics. It is only activated Port statistics

when action is already acutely re- Overview Netload m/s Discards Error

quired. This is why it is advisable

Port 1 Port 2 Port 3
to permanently monitor the log- 0 387 0
Partner: 10.1.9.2 Partner: 10.1.9.29 Partner: 10.1.9.15
ical PROFINET communication.
Not only does it report when Port 4 Port 5 Port 6
0 1215 0
events occur that are externally Partner: 10.1.9.23 Partner: 10.1.9.16 Partner: 10.1.9.22

not even visible, it also keeps the

Port 7 Port 8 Port 9
historical diagnostic data ready 0
Partner: 10.1.9.24
0%
Partner: -
0%
Partner: -

to recognise trends and identify

the causes of events.

Indu-Sol 9 White paper

2
SEARCH AND YOU WILL
FIND!
Selected causes of unstable network
communication

The choice of infrastructure components plays an essen-

tial role here, as managed switches do not just determine
port statistics, but also specify
Port statistics
them with regard to the present-
Overview Netload m/s Discards Error

ed quality parameters (see also

Chapter 3). Port 1 Port 2 Port 3
0 0 0
Partner: 10.1.9.2 Partner: 10.1.9.29 Partner: 10.1.9.15
Electromagnetic influences on
the data communication can also
Port 4 Port 5 Port 6
be detected in different ways. The 0
Partner: 10.1.9.23
0
Partner: 10.1.9.16
15843
Partner: 10.1.9.22

measurement-based checking of
Port 7 Port 8 Port 9
these leakage currents poses two 0 0% 0%
Partner: 10.1.9.24 Partner: - Partner: -
challenges: On the one hand, in a
network there are typically many
lines reaching several kilometres in total length. Moni- Even faulty telegrams (error telegrams) − caused perhaps by defective cables or
electromagnetic couplings − are displayed in the web interface.
toring these lines with a mobile measuring device (for ex-
ample, a leakage current clamp) would be difficult, and
could only be done one line after the other. On the other rarily. In order to avoid the necessity of regular manual
hand, the load on the equipotential bonding and the data measurements so that the level of the shield currents can
shielding is not constant. It changes for certain work pro- at least be approximated over time, permanent monitoring
cesses, for example, or if devices are switched on tempo- is recommended, including alarms as soon as measured

Indu-Sol 10 White paper

values exceed a critical threshold. The solution is to in- Statistically seen, the most common
tegrate such a mechanism into the switch (see leakage causes of a deterioration in PROFINET
current display below). communication quality are an out-of-
date network infrastructure, electro-
However, only managed switches make available the magnetic couplings, or an excessive
diagnostic data they determine. And even within these network limit. This can all be deter-
managed switches there are differences in functionali- mined through permanent monitoring
ty. This means that switches should always be selected of the PROFINET network.
while keeping their purpose in mind (see “Quality of
Service” in Chapter 3).

Leakage current Time basis History

Last minute
Peak RMS Last hour

19 mA

0 mA 200 mA 400 mA

Indu-Sol 11 White paper

3
THE SWITCH AS CENTREPIECE
OF NETWORKED STRUCTURES

Communication in PROFINET and Ethernet networks in determining how the network performs. Planners
is controlled by intelligent infrastructure components and designers can thus greatly influence the stability
(switches). These connect the devices in a PROFINET/ of the later application's communication by selecting
Ethernet network, read out incoming data packages, switches that at a minimum satisfy requirements.
and transmit them to the port with the right receiver.
The packages are also checked for errors and discard-
ed if necessary. Such and similar events are docu-
mented in the port statistics that each switch keeps.
PROFINET switches also give PROFINET data priority
over acyclical non-PROFINET data.
Beyond their role as data intersections (comparable
to logistics centres), switches can contribute signif-
icantly to efficient network diagnostics. The perfor-
mance capacity of the switches is therefore significant Source: Scene from “Switches − Professionals in the network | Technology in simple terms”

Unmanaged vs. managed

switches
Switches can be basically divided into two groups: Un- be integrated in the hardware configuration either: If the
managed and managed switches. On cost grounds, unman- device fails, the PLC does not receive a diagnostic message.
aged switches are used in many machines and systems. As the name already suggests, the use of managed
They are relatively cheap to buy but only provide a limited switches makes network management much easier. On
range of functions. For example, the port statistics they the one hand, they offer extensive possibilities for de-
determine cannot be retrieved and are therefore worthless. vice and port configuration, as well as for the activation
Unmanaged switches are not detected during a topology of additional functions such as port mirroring, bandwidth
scan and are therefore not included in the documenta- control, or redundancy support. On the other hand, they
tion. They can sometimes even negatively affect network supply important information for diagnostics, as already
communication by not supporting certain protocol-typical shown in Chapter 2 (errors, discards, network load, leak-
functions (e.g. for PROFINET). Unmanaged switches cannot age current).

Indu-Sol 12 White paper

 This diagnostic data is a basis for the daily work of “More networked than you thought
commissioning specialists, operators and service tech- – because of the simple connection
nicians. If the transmission of this diagnostic data is possibilities in Ethernet-based net-
combined with an alarm function when critical values works, their expansion can often only
are exceeded, the persons responsible can be specif- be really captured by a topology scan.
ically warned as soon as the first signs of worsening The extent of networking of your own
communication appear. This makes it possible to plan machine or system is usually grossly
maintenance measures according to needs. In princi- underestimated.”
ple, only managed switches provide the port statis-
tics (and thus the basis for diagnostics), so that they
can be queried via SNMP. The status of the network
can then be read off at any time using the diagnostic
data. Furthermore, only managed switches are detected network management considerably thanks to a wide
during a topology scan, which is however essential to variety of configuration options (e.g. IP address, name,
gain a complete overview of the network. They simplify additional functions such as port mirroring).

Only managed switches are detected during a topology scan (see red marking in image). However, this is absolutely necessary to gain a complete
overview of the network, and to enable a precise reaction if an error occurs.

Indu-Sol 13 White paper

3
THE SWITCH AS CENTREPIECE
OF NETWORKED STRUCTURES

Other important switch

properties
CONFORMANCE CLASS NETLOAD CLASS
The range of functions of PROFINET IO devices is clearly Netload Classes define the robustness of a device in
divided into Conformance Classes. These offer a practical relation to the network traffic to be processed. Devices
summary of various minimum properties. Conformance are categorised into Netload Class I to III, with Class III
Class B is of special importance for infrastructure compo- ensuring the highest possible network availability.
nents such as switches, as it contains the basic PROFINET
functionalities, thus covering the requirements for 99% of
the PROFINET networks installed in the field. Type Netload Class I Netload Class II Netload Class III

Unicast < 0.1% < 1% < 10%

Multicast/
Conformance Class C < 1% < 5% < 10%
Broadcast
Clock-synchronised communication
for PROFINET IRT

Conformance Class B
Cyclical and acyclical data exchange;
Device diagnostics;
Continuous determination of topology;
Network diagnostics;
Device identification;
PI-certification

Conformance Class A
IT-compatible switches
incl. prioritisation

Indu-Sol 14 White paper

 QUALITY OF SERVICE (QoS) The properties of the switches installed
The performance a switch really has to deliver depends in the network significantly influence
on the application for which it is intended. Access layer network performance. In order to lay
switches (100 Mbps, e.g. PROmesh P9) generally suffice the basis for stable PROFINET data
for machine controllers in industrial automation. However, communication, we recommend using
there may be even here applications that require a gi- fully managed switches with Conform-
gabit-capable distribution layer switch that also allows ance Class B and Netload Class III cer-
different connection types such as copper or optical fibres tification. The performance capacity of
(e.g. PROmesh P20). A distribution layer switch is nec- the switches must be at least as high
essary at the latest for communication across the entire as the Quality of Service requirements
network, where individual connections or ports should be of the applications communicating in
specifically permitted or blocked. Core layer switches are the network.
used at office level, where high data transmission rates
and highly redundant forwarding are required. The illus-
tration on page 18 shows which switch makes sense at
which level. Further information on networks structures is
also provided.

Watch the explanatory video for a concise summary

of the most important switch functions:
www.indu-sol.com/pn19-switch

Indu-Sol 15 White paper

4
CENTRAL CONDITION
MONITORING BY SOFTWARE
Bringing together distributed diagnostic data

In order to receive an overview of the current “health”

of the network at the push of a button, it is best to in-
stall network monitoring software centrally on a server
including the corresponding database. The software then
cyclically retrieves the diagnostic data from the managed
switches in the network, and bundles this into an entire
status. If required, a scaled analysis up to the individual
You surely know somebody with a huge library at switch port is then possible:
home. The names of famous authors stare down from the
shelves, putting the books' owner in an intellectual light. Netload per port

Even so, few of the pages in these shelves are dog-eared.

100
The written contents have not yet been so much as pe-

ms
90

rused. 80

70
The same is true for many industrial companies. The
Percent %

60

desire for data is constantly growing. An increasing 50

40
number of sensors records operating parameters such as
30

temperature, air humidity, speeds, vibrations, and so on. 20

Nonetheless, the mountain of data is rising quicker than 10

0
1 2 3 4 5 6 7 8 9
it can be processed by the human operators (Big Data). 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200
Port load Total utilization
Besides, the job of a maintenance employee usually in-
volves more than just evaluating data. Algorithms have
to be created that transform the data into information This current and historical status data forms an indis-
(Smart Data), while also detecting and reporting anom- pensable basis for network management. The basic task
alies. of network monitoring is to visualise all deviations when
In the context of industrial networks, the challenge monitoring processes, and thus to detect network anom-
is to centrally bundle into a clear overall picture, the alies or overloads in the form of discards and jitters, for
port statistics collected decentrally by the switches. example. If this monitoring is automated and combined
Here, two advantages of using managed switches become with alarm management that issues notifications when
clear: They are the only ones that make diagnostic data anomalies occur, the entire package remains manageable
available for retrieval. This data is also captured by a in terms of the work to be performed by technical staff.
topology scan so that anomalies are quickly located if re- (See also www.indu-sol.com/pn19-monitoring).
quired. This saves valuable time during troubleshooting With this in mind, it is difficult to understand why
and therefore money. network monitoring in industrial automation has been

Indu-Sol 16 White paper

mostly relegated to the margins until now. In today's The requirements on the availability
IT, not a single network is commissioned without and security of networks are increas-
monitoring and a responsible administrator. This basic ing. This is due to global competition
approach is often forgotten, however, at the factory and the fact that automation networks
floor switch. This occurs in spite of the fact that the are opening more to the outside. Both
highest requirements are placed on network availabili- factors can only be addressed by a
ty in industrial automation, as well as on the punctual monitoring system that will make avail-
transmission of time-critical data. Hard cash is lost able current and historical diagnostic
with every wasted minute. This is in stark contrast to data where required. This system must
IT, where a temporary network failure, or a print job also be able to alert operators to anom-
or email that is a few minutes late, are generally un- alies that it detects.
important. In particular, companies that require their
automated applications around the clock depend on
the availability of networks for machine and system
control, as unplanned downtimes cannot be made up
for.
As well as securing availability, network monitoring
has a second important function: It recognises anom-
alies with security relevance. This often begins with
a supposedly trivial matter:
Thorough knowledge of your
own network. In order to even
detect unknown devices that
suddenly communicate in the
network, you must first know
who “innately” belongs to
the network. Only if the com-
munication in the network is
monitored can programming
access to the controller be
detected, for example, and
evaluated as to whether this Figure: The fully managed switches monitor the data communication quality decentrally, and make this data available to a
central evaluation unit. If required, this shows the current network status or sends alarms in the case of anomalies.
process is authorized.

Indu-Sol 17 White paper

5
CONCLUDING THOUGHTS ON
A SUSTAINABLE NETWORK
STRUCTURE
Networks are increasingly becoming a critical asset ries out diagnostics. The selection of the infrastructure
for today's machines and plants. The level of automa- components according to the required performance, and
tion in manufacturing companies is increasing all the their integration into a future-proof network structure,
time. As a result, the continuous availability and high are the basis which should be given special attention
security of these networks has also gained in impor- even in the planning phase. Of course, there is no one
tance. Also rising constantly is the demand for data, solution that fits all, or even a template, as individual
as this forms the basis for network management, and requirements must always be taken into account.
provides important starting points for network diagnos- At present there are two methods for counteracting
tics, when required. network overload:
The growing data volume also raises the requirements 1. TCP/IP communication and real-time communica-
on technology: High data throughput rates, quick reac- tion are separated inside the machine or plant.
tion times, several protocols communicating over the 2. The bandwidth in the machine network is expanded
same infrastructure. Without targeted monitoring of by integrating gigabit infrastructure components.
the network, it is almost impossible today to ensure its Both approaches seem expedient at first because they
availability and security. However, monitoring only car- fulfil the current performance requirements. However, it
IT

H
C
IT
SW

1 Gbps 10 Gbps
PROmesh P28-RL
IIT

2x 1 Gbps 2x 1 Gbps
1 Gbps

PROmesh P20 PROmesh P20 PROmesh P20

100 mbps 100 mbps
100 mbps
OT

PROmesh P9 PROmesh P9 PROmesh P9

PROFINET-INspektor® NT PROmesh P9
PROmesh P9

PROmesh FX PROmesh FX
PROmesh P7X

Indu-Sol 18 White paper

is clear that in the long term these solutions will not be
able to keep up with the rapid growth in data traffic,
and the increasing degree of networking at OT level.
The result of many discussions on this topic with
? THE QUICK-CHECK FOR
STABLE AND SECURE
NETWORK OPERATION
IN PROFINET ‒ MAKE
SURE TO...
users from various production sectors makes apparent
that between OT and IT an additional, high-perform- ☑ Check performance requirements
of later applications on the net-
ing and powerful communication level has already been work (data throughput, availabili-
created or is being introduced: With this Industrial ty, bandwidth, …)

☑ Plan
Information Technology (IIT) level, the bulk of TCP/IP network structure
data traffic and the machine/machine communication (PROFINET vs. non-PROFINET
can be processed. The actual machine network (OT traffic or connection of acyclically
communicating devices)
level) is relieved considerably and gains reserves so
⇨ Consider the PROFINET
that it can meet the increasing demands on machine Design and Commissioning
control. Guideline from PI

☑ Select suitable infrastructure com-

At every level, switches that are adapted to indi-
vidual performance needs are also used (see “Quality ponents

☑ Simulate draft with a planning

of Service” in Chapter 3, page 13). In the OT area,
PROFINET is still only used, whereby access layer tool and check for practicality
(network access points / measuring
switches are sufficient. On the other hand, the IIT area
points, network load, security
requires gigabit-capable switches that are equipped precautions, …)

☑ Select
with high-performance quality of service functions. The
and set up monitoring
switches in the IT area are designed for high transmis- system

!
☑ Specify
sion rates and highly redundant transmission.
responsibilities for net-
The logical consequence of this structuring would work management; if necessary
then be to define at least one fixed administrator for train/instruct personnel

☑ Perform
the IIT and OT networks, who would bring together acceptance
knowledge from network technology that is typically measurement with documentation
available in the IT area and the requirements of OT (current topology plan + report
on results of quality analysis)
networks (high availability, shortest possible delays).
Alternatively, the administration can be managed in
collaboration with external service providers. It will
certainly be interesting to see which solutions are ulti-
mately successful on the market.

Indu-Sol 19 White paper

 PROFINET - The film
| Technology in simple terms
www.indu-sol.com/pn19-film
or simply use the QR code

Indu-Sol GmbH
Blumenstrasse 3
04626 Schmoelln
Telephone: +49 (0) 34491 5818-0
Telefax: +49 (0) 34491 5818-99
[email protected] SWITCHES - Professionals in the network
| Technology in simple terms
www.indu-sol.com
www.indu-sol.com/pn19-switch
Certified according to DIN EN ISO 9001:2015 or simply use the QR code
Updated 11/2019