SYNOPSISracluster [-f conf] [-m agr(s)] [-M mode(s)] [raoptions]
Racluster reads argus data from an argus-data source, and clusters/merges the records based on the flow key criteria specified either on the command line, or in a racluster configuration file, and outputs a valid argus-stream. This tool is primarily used for data mining, data management and report generation.
The default action is to merge status records from the same flow and argus probe, providing in some cases huge data reduction with limited loss of flow information. Racluster provides the ability to modify the flow model key, either using the "-m" option, or in the racluster.conf file, allowing records to be clustered based on any number of attributes. This supports the development of important reports, such as MPLS LSP usage statistics, DiffServe flow marking policy verification, VLAN group behavior, IP distance related measurements, routing loop detection, traceroute path data recovery, and complex availability/reachability reports, to name just a few useful applications.
Please see racluster.5 for detailed information regarding racluster configuration.
OPTIONSRacluster, like all ra based clients, supports a number of ra options including filtering of input argus records through a terminating filter expression, and the ability to specify the output style, format and contents for printing data. See ra(1) for a complete description of ra options. racluster(1) specific options are:
- -m aggregation object
- Supported aggregation objects are:
- use a null flow key.
- argus source identifier.
- source mac(ether) addr.
- destination mac(ether) addr.
- source mpls label.
- destination label addr.
- source vlan label.
- destination vlan addr.
- source IP addr/[cidr len | m.a.s.k].
- destination IP addr/[cidr len | m.a.s.k].
- sorted src and dst IP addr/cidr len.
- transaction protocol.
- source port number.
- destination port number.
- source TOS byte value.
- destination TOS byte value.
- src -> dst TTL value.
- dst -> src TTL value.
- src -> dst TCP base sequence number.
- dst -> src TCP base sequence number.
- intermediate node, source of ICMP mapped events.
- -M modes
- Supported modes are:
- Do not generate an aggregate statistic for each flow. This is used primarily when the output represents a single object. Primarily used when merging status records to generate single flows that represent single transactions.
- Generate data suitable for producing RMON types of metrics.
- Process each input file independantly, so that after the end of each inputfile, racluster flushes its output.
- Replace each inputfile contents, with the aggregated output.
Verbose operation, printing a line of output for each input file processed.
Very useful when using the ra() -R option.
INVOCATIONA sample invocation of racluster(1). This call reads argus(8) data from inputfile and aggregates the IP protocol based argus(8) data, based on the source and destination address matrix and the destination port used by tcp flows, and report the metrics as a percent of the total. For most services, this provides service specific metrics on a client/server basis.
- racluster -% -r inputfile -m saddr daddr dport - \ tcp and syn and synack This call reads argus(8) data from inputfile and generates the path information that traceroute use would generate (assuming that traceroute was run during the observation period).
- racluster -r inputfile -m saddr daddr sttl inode -w - - icmpmap | \ rasort -m sttl -s saddr dir daddr inode avgdur spkts
COPYRIGHTCopyright (c) 2000-2008 QoSient. All rights reserved.
Carter Bullard ([email protected]).