User Manual for NOP1840 PEP-Box Server
Features
Newtec proprietary
Version 1.0
Confidentiality :
Unrestricted
Page 46 / 96
The assignment of the guaranteed bandwidth is independent from the distribution of the remaining
bandwidth. Therefore if two rules receive the same share of the remaining bandwidth but the
guaranteed bandwidth of one rule is higher, the rule with the higher guaranteed bandwidth will receive
a bigger share of the overall bandwidth.
E. g. 1.5 Mbit/s are shared by rule_01 and rule_02. The guaranteed bandwidth of the rules is 400
kbit/s for rule_01 and 100 kbit/s for rule_02. The guaranteed bandwidth is assigned to the rules. The
remaining part of the overall bandwidth is 1 Mbit/s (1.5 Mbit/s – 400 kbit/s – 100 kbit/s). This is then
split equally between the rules, each rule receiving 500 kbit (1 Mbit/2). As a result, rule_01 will receive
900 kbit of the overall available bandwidth and rule_02 will receive 600 kbit/s.
The remaining available bandwidth is not necessarily distributed equally among active rules but can be
allocated to rules in a weighted manner. For each rule a weighting factor can be configured that
decides on the proportion of the bandwidth allocated to this rule. The weighting factor is by default the
peak bandwidth assigned to the rule. That means that rules with a higher configured peak bandwidth
will receive a higher share of the available bandwidth. This can be altered by configuration of a
weighted bandwidth. If available, the weighted bandwidth replaces the peak bandwidth as weighing
factor for the distribution of the remaining bandwidth.
E. g. if two rules have a peak bandwidth of 1 Mbit/s, both rules will receive the same share of the
available remaining bandwidth. But if a weighted bandwidth of 500 kbit/s is additionally configured for
one of the rules, this rule with then receive only half as much of the available remaining bandwidth as
the rule without the configured weighted bandwidth.
5.4.2.6
Queuing
If the bandwidth currently assigned to a rule is not sufficient to forward all currently captured traffic, the
NOP1840 will queue incoming data packets per rule. In case the configurable maximum queue size for
a rule is exceeded, packets in the queue are dropped. A priority can be assigned to data packets with
a specified source or target IP address or a specified TOS value. Packets with a lower priority are
dropped prior to packets with a higher priority. Queued packets with a higher priority are also sent prior
to packets with a low priority.
5.4.2.7
Hierarchical Traffic Policing
The traffic policing mechanisms described above can not only be applied to rules but also to groups of
rules. A guaranteed bandwidth, a maximum peak bandwidth and a weighting factor for bandwidth
distribution can be assigned to a group.
In case a rule belongs to a group, shaping rules defined for the rule apply as well as shaping rules
defined for the group. E. g. if two rules with a peak bandwidth of 1 Mbit/s are assigned to a group with
a peak bandwidth of 1.5 Mbit/s, and both rules are active, none of the two rules will reach the peak
bandwidth of 1 Mbit/s, but, in case no weighting factor or guaranteed bandwidth is configured, both
rules are assigned half of the peak bandwidth of the group, i.e. 750 kbit/s.
5.4.2.8
Accounting
Accounting information about the traffic assigned to rules, e.g. terminals, and groups is written into an
accounting file at the NOP1840. The CSU format allows easy processing of the accounting data for all
shaped traffic.
