On 03/02/2015 11:28, nfdump-discuss-requ...@lists.sourceforge.net wrote:
> In a wireshark capture I cant seem to see any field which would 
> indicate the amount of bytes?  I see initiator octets and responder 
> octets change but I don't know what these fields are used for.
>
Periodic byte counters were added as a feature in 8.4(5) and 9.1(2):
http://www.cisco.com/c/en/us/td/docs/security/asa/roadmap/asa_new_features.html

I think the initiator/responder octets are indeed those counters. 
Example I see here with ASA 8.4(7):

     FlowSet 1
         FlowSet Id: (Data) (263)
         FlowSet Length: 440
         Flow 1
             Flow Id: 1877861196
             SrcAddr: x.x.x.140 (x.x.x.140)
             SrcPort: 33755
             InputInt: 2
             DstAddr: y.y.y.10 (y.y.y.10)
             DstPort: 80
             OutputInt: 11
             Protocol: 6
             IPv4 ICMP Type: 0
             IPv4 ICMP Code: 0
             Post NAT Source IPv4 Address: x.x.x.140 (x.x.x.140)
             Post NAT Destination IPv4 Address: 192.168.10.104 
(192.168.10.104)
             Post NAPT Source Transport Port: 33755
             Post NAPT Destination Transport Port: 80
             Firewall Event: Unknown (5)
             Extended firewall event code: ignore (0)
             Observation Time Milliseconds: Feb  3, 2015 
11:43:05.127000000 GMT
             Initiator Octets: 838
             Responder Octets: 9284
             StartTime: Feb  3, 2015 11:42:03.935000000 GMT

I got the above using

     tshark -ieth0 -nnV -s0 -d udp.port==9001,cflow udp port 9001

Note: this won't show anything of interest until after the flow template 
message has been read. By setting

     flow-export template timeout-rate 1

you won't have to wait more than one minute before the "no template 
found" messages disappear.

I also have:

     flow-export delay flow-create 10

in an attempt to aggregate flow data a bit.

I do get plausible flows (i.e. the "In Byte" and "Out Byte" columns in 
nfsen are not all the same). So it seems to me you ought to capture some 
tshark data, correlate this with the nfdump output at the same time, and 
see if the problem is the data from the ASA or the way it's recorded by 
nfcapd and displayed by nfdump.

Note: I believe I built nfdump with:
./configure --enable-nfprofile --enable-nftrack --enable-nsel

$ nfdump -V
nfdump: Version: NSEL-NEL1.6.12 $Date: 2014-04-02 20:08:48 +0200 (Wed, 
02 Apr 2014) $

HTH,

Brian.


------------------------------------------------------------------------------
Dive into the World of Parallel Programming. The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net/
_______________________________________________
Nfdump-discuss mailing list
Nfdump-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfdump-discuss

Reply via email to