Wednesday, February 27, 2013

[sFlow] Use sflowtool to parse sFlow datagram

In order to test and understand sFlow more in details, I prepare the following environment below. Switch 1 and 2 are emulated using Open vSwitch. In the previous sFlow article, there is a sFlow setting on Open vSwitch. Please check it out.

            192.168.12.201           10.3.207.244           192.168.12.202
            +------------+        +-----------------+       +------------+
            |  PC1       |        | sFlow Collector |       |  PC2       |
            |            |        |                 |       |            |
            |            |        |                 |       |            |
            +-----+------+        +---^--------^----+       +-----+------+
                  |                   |        |                  |
                  |                   |        |                  |
                  |                   ^ sFlow  ^                  |
                  +---------+         | Data   |       +----------+
                            |         ^        ^       |
                            |         |        |       |
                            |         ^        ^       |
                            |         |        |       |
                   +--------+---------++      ++-------++---------+
                   |    Switch 1       |      |   Switch 2        |
                   | 10.3.207.142      +------+ 10.3.207.143      |
                   +-------------------+      +-------------------+

sFlowtool is an open source toolkit for us to leverage the functionality of parsing sFlow datagram. In my environment, when I ping PC2 from PC1, then Switch 1 and 2 will send sFlow Counter and Flow Sample datagram to the sFlow collector. Now, I will use sFlowtool to parse  the information that it gets.

> ./sflowtool
Counter Sample:startDatagram =================================
datagramSourceIP 0.0.0.0
datagramSize 144
unixSecondsUTC 1359534209
datagramVersion 5
agentSubId 0
agent 10.3.207.142
packetSequenceNo 403
sysUpTime 947000
samplesInPacket 1
startSample ----------------------
sampleType_tag 0:2
sampleType COUNTERSSAMPLE
sampleSequenceNo 95
sourceId 0:4
counterBlock_tag 0:1
ifIndex 4
networkType 6
ifSpeed 1000000000
ifDirection 1
ifStatus 3
ifInOctets 126361
ifInUcastPkts 1072
ifInMulticastPkts 0
ifInBroadcastPkts 4294967295
ifInDiscards 0
ifInErrors 0
ifInUnknownProtos 4294967295
ifOutOctets 137350
ifOutUcastPkts 1135
ifOutMulticastPkts 4294967295
ifOutBroadcastPkts 4294967295
ifOutDiscards 0
ifOutErrors 0
ifPromiscuousMode 0
endSample   ----------------------
endDatagram   =================================
startDatagram =================================
datagramSourceIP 0.0.0.0
datagramSize 144
unixSecondsUTC 1359534210
datagramVersion 5
agentSubId 0
agent 10.3.207.244
packetSequenceNo 404
sysUpTime 948000
samplesInPacket 1
startSample ----------------------
sampleType_tag 0:2
sampleType COUNTERSSAMPLE
sampleSequenceNo 95
sourceId 0:5
counterBlock_tag 0:1
ifIndex 5
networkType 6
ifSpeed 100000000
ifDirection 2
ifStatus 1
ifInOctets 0
ifInUcastPkts 0
ifInMulticastPkts 0
ifInBroadcastPkts 4294967295
ifInDiscards 0
ifInErrors 0
ifInUnknownProtos 4294967295
ifOutOctets 0
ifOutUcastPkts 0
ifOutMulticastPkts 4294967295
ifOutBroadcastPkts 4294967295
ifOutDiscards 0
ifOutErrors 0
ifPromiscuousMode 0
endSample   ----------------------
endDatagram   =================================

Flow Sample:startDatagram =================================
datagramSourceIP 10.3.207.244
datagramSize 216
unixSecondsUTC 1359597631
datagramVersion 5
agentSubId 0
agent 10.3.207.142
packetSequenceNo 941
sysUpTime 2174000
samplesInPacket 1
startSample ----------------------
sampleType_tag 0:1
sampleType FLOWSAMPLE
sampleSequenceNo 72
sourceId 0:4
meanSkipCount 64
samplePool 4010
dropEvents 0
inputPort 4
outputPort 3
flowBlock_tag 0:1001
extendedType SWITCH
in_vlan 0
in_priority 0
out_vlan 0
out_priority 0
flowBlock_tag 0:1
flowSampleType HEADER
headerProtocol 1
sampledPacketSize 102
strippedBytes 4
headerLen 98
headerBytes 00-AB-77-E3-4B-00-00-AB-71-13-2D-00-08-00-45-00-00-54-00-00-40-00-40-01-9F-C5-C0-A8-0C-CA-C0-A8-0C-C9-08-00-C4-0E-55-06-08-57-28-96-AD-FD-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00
dstMAC 00ab77e34b00
srcMAC 00ab71132d00
IPSize 84
ip.tot_len 84
srcIP 192.168.12.202
dstIP 192.168.12.201
IPProtocol 1
IPTOS 0
IPTTL 64
ICMPType 8
ICMPCode 0
endSample   ----------------------
endDatagram   =================================

If you use this, then you will get another format of data:
> ./sflowtool -l
Counter Sample:
CNTR,10.3.207.142,3,6,1000000000,1,3,112616,955,1,4294967295,0,0,4294967295,123043,1012,4294967295,4294967295,0,0,0

Flow Sample:
FLOW,10.3.207.142,4,3,00ab71132d00,00ab77e34b00,0x0800,0,0,192.168.12.202,192.168.12.201,1,0x00,64,0,0,0x00,102,84,64


Monday, February 25, 2013

[C++ STL] Containers

We all know that using data structure well is very important in programming, because it affects the performance, data accuracy, maintenance no matter what kind of the program you write. In C language, we have to provide our data structure by ourselves or by searching for related library and grab it to use. But, if possible ( I mean if your program is able to use g++ compiler and the environment ), you can consider to just use Containers (C++ STL) in your program. It is useful and powerful. For more information, here is the official web site : http://www.cplusplus.com/reference/stl/

Container class templates

Sequence containers:

Container adaptors:

Associative containers: