Hello Peter,
 
I have a script that every 5 minutes runs the following command:
 
"nfdump -r /opt/data/netflow/live/nfcapd.time_stamp -n 5 -s dstip:p/pps
-l 1000"
 
Most of the times the output is fine but from time to time I get a
record like this in the list:
 
Date first seen          Duration Proto      Dst IP Addr    Flows
Packets    Bytes      pps      bps   bpp
 
291
 
290
 
282
2009-09-08 15:41:50.129 4295023.076 TCP        xx.xx.xx.xx     5056
13858   673426        0        1    48
 
228
 
225
 

I looked at all the flows for that particular IP and found this
particular one:
 
Date flow start          Duration Proto      Src IP Addr:Port
Dst IP Addr:Port   Packets    Bytes Flows
2009-09-08 15:42:46.333 4294966.872 TCP     187.47.192.161:49831 ->
xx.xx.xx.xx:14823        3      144     1
Summary: total flows: 1, total bytes: 144, total packets: 3, avg bps: 0,
avg pps: 0, avg bpp: 48
Time window: 2009-09-08 15:11:56 - 2009-10-28 08:45:33
Total flows processed: 1438830, Records skipped: 0, Bytes read: 74820240
Sys: 0.076s flows/second: 18689016.5 Wall: 0.076s flows/second:
18909083.7
 
Or in raw format:
 
Flow Record:
  Flags       =       0x00000000
  size        =               52
  mark        =                0
  srcaddr     =   187.47.192.161
  dstaddr     =      xx.xx.xx.xx
  first       =       1252438966 [2009-09-08 15:42:46]
  last        =       1256733933 [2009-10-28 08:45:33]
  msec_first  =              333
  msec_last   =              205
  dir         =                0
  tcp_flags   =             0x 2 ....S.
  prot        =                6
  tos         =                0
  input       =              188
  output      =                0
  srcas       =                0
  dstas       =                0
  srcport     =            49831
  dstport     =            14823
  dPkts       =                3
  dOctets     =              144
 
Summary: total flows: 1, total bytes: 144, total packets: 3, avg bps: 0,
avg pps: 0, avg bpp: 48
Time window: 2009-09-08 15:11:56 - 2009-10-28 08:45:33
Total flows processed: 1438830, Records skipped: 0, Bytes read: 74820240
Sys: 0.072s flows/second: 19713240.5 Wall: 0.071s flows/second:
20115619.1
 
 
 
I think that the stats are good and that IP should be in the top 5 list
but for some reason that particular flow messes the output up.

However I seems that it generates the right stats but for some reason it
doesn't display it right. Or it uses a different algorithm to generate
the stats and another algorithm for the output.
 
I this a known issue? is there a workaround? or a way to filter out
these "invalid" records (flows older than the default aging time)?

I'm using:
nfdump: Version: 1.5.7
nfsen-1.3 

Thank you,
Bogdan.





------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with 
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Nfsen-discuss mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/nfsen-discuss

Reply via email to