Hi Gerhard,

I suggest as first step to upgrade to a recent 0.14 release and
compile it yourself. Since you speak of 70GB traffic, make sure
when configuring the package for compilation the --enable-64bit
switch is specified.

About 1) i can suggest to turn on some internal buffering (at this
propo see the email i've just sent):

http://www.mail-archive.com/[email protected]/msg01978.html

About 2) i can't provide an answer. Trusting your awk is correct
(which absolutely looks like it is) it might be a bug fixed in
due course?

Cheers,
Paolo

On Thu, Jun 14, 2012 at 11:27:46AM +0100, Gerhard Lazu wrote:
> I have pmacct v0.11.6 running on Ubuntu 10.04 for about 8 hours now, using
> the following config:
> 
> daemonize: true
> pidfile: /var/run/pmacctd.pid
> syslog: daemon
> !
> ! interested in in and outbound traffic
> aggregate: src_host,dst_host,src_port
> ! on this interface
> interface: eth0
> 
> When I sum the counters for port 6201, I get:
> 
> $ pmacct -c src_port -N 6201 -n bytes | awk '{ print $1 " Bytes"; print $1
> / 2^30 " Gigabytes" }'
> 709264151 Bytes
> 0.660554 Gigabytes
> 
> When I do the summing in awk, I get:
> 
> $ pmacct -s | awk '/ 6201 / { total+=$5 } END { print total " Bytes"; print
> total / 2^30 " Gigabytes" }'
> 13595724124 Bytes
> 12.662 Gigabytes
> 
> However, I know for a fact that the service running on port 6201 has
> transferred at least 70 Gigabytes within this period, all TCP traffic.
> 
> This leads me to 2 questions:
> 
> 1. Why are the pmacct counters off by that much?
> 
> 2. Why the huge difference between pmacct -c & pmacct -s with awk summing?
> 
> Cheers, Gerhard
> 
> ------------------------------
> Twitter <http://twitter.com/#!/gerhardlazu> Github<https://github.com/gerhard>
>  Blog <http://gerhardlazu.com/>

> _______________________________________________
> pmacct-discussion mailing list
> http://www.pmacct.net/#mailinglists

_______________________________________________
pmacct-discussion mailing list
http://www.pmacct.net/#mailinglists

Reply via email to