This README is aimed at explaining the use of programs Filter and Aggprefs. Filter is used to measure the filtering efficiency of ASs in a network-wide implementation of DRAGON. Aggprefs is used for determining the set of aggregation prefixes that the ASs would generate under a network-wide implementation of DRAGON. Their optimal use is in a UNIX-based operating system such as Linux.
In order to execute either of these programs, it might be necessary to initially re-compile them. To do so, simply open a terminal in the program's main directory and type:
make clean
make
To execute the programs from a terminal in their main directory, simply type:
- For Filter: ./bin/filter <NETWORK_FILE> <PREFIX_FILE>
- For Aggprefs: ./bin/aggpregs <NETWORK_FILE> <PREFIX_FILE>
The arguments <NETWORK_FILE> and <PREFIX_FILE> are described in further detail in the INPUTS text file. Both arguments are mandatory.
The program Filter outputs a file named "filtering_efficiency.txt" with some statistics regarding the filtering efficiency when considering a network-wide deployment of DRAGON. The output file contains the following sections:
- Filtering Efficiency per AS
Each of the first lines in the file has the format AS [(STUB)]: <ORIGINAL_#FIB_ENTRIES> -> <NEW#_FIB_ENTRIES>, <FILTERING_EFFICIENCY>
- AS [(STUB)] is the AS to which the data refers. In the event that this AS is a stub, then the indication (STUB) is included.
- <ORIGINAL_#_FIB_ENTRIES> is the amount of entries in this AS's FIB after all routes were propagated, without the use of DRAGON
- <NEW_#_FIB_ENTRIES> is the amount of entries in this AS's FIB after all routes were propagated, once DRAGON is implemented
- <FILTERING_EFFICIENCY> is the measure of DRAGON's filtering efficiency in this AS, calculated as <NEW_#FIB_ENTRIES> / <ORIGINAL#_FIB_ENTRIES>
The final two lines of this section contain the maximum and the minimum filtering efficiencies obtained, respectively.
- Complementary Cumulative Distribution of Filtering Efficiency
The following section contains pairs of values which are used to graph the complementary cumulative distribution function of filtering efficiency. In each pair , is the number of ASs which have a filtering efficiency that is either equal to, or greather than . This section is divided in two sub-sections: the first one pertains to all ASs, while the second one pertains exclusively to non-stub ASs.
The program Aggprefs outputs a prefix file "prefixes_with_agg.txt" with the same format as the input <PREFIX_FILE> which contains all of the IPv4 prefixes received in the input file, as well as the newly generated aggregation prefixes.