Hey all, having an issue with a pmacct setup that seems a bit confusing
to me.  Didn't see anything in the mailing list archives that I think
relates to this situation.

Basically what's happening is that we have two pmacct setups.

1 - a mirror of all traffic on our class C.  This is the number we know
is right.  The setup there looks like this:

debug:false
interface:eth1

aggregate[in]:dst_host
aggregate[out]:src_host
aggregate_filter[in]:  dst net 64.x.x.0/24 and not src net 64.x.x.0/24
aggregate_filter[out]: src net 64.x.x.0/24 and not dst net 64.x.x.0/24
plugins:pgsql[in], pgsql[out]
sql_data:frontend
sql_db:dbname
sql_table:tablename
sql_user:user
sql_passwd:password
sql_refresh_time:600
sql_optimize_clauses:true
sql_history:1d
sql_host:localhost
sql_recovery_logfile:/home/xx/logs/recover-daily.log
pidfile:/home/xx/logs/pmacctd-daily.pid

sql_history_roundoff:1d
plugin_pipe_size: 512000
plugin_buffer_size: 4096
! next must be prime....
sql_cache_entries: 99991


2 - We also have a firewall out in the field which has a couple of IPs from
the class C assigned to it.  In theory those IPs bandwidth should match
almost exactly the data from setup #1.  The second servers setup looks
like this:

debug:false
interface:eth3

aggregate[in]:dst_host,dst_mac
aggregate[out]:src_host,src_mac
aggregate_filter[in]:  dst net (192.168.x.0/23 or 64.x.x.0/24) and not src net 
(192.168.x.0/23 or 64.x.x.0/24)
aggregate_filter[out]: src net (192.168.x.0/23 or 64.x.x.0/24) and not dst net 
(192.168.x.0/23 or 64.x.x.0/24)
plugins:pgsql[in], pgsql[out]
sql_data:frontend
sql_db:dbname
sql_table:tablename
sql_user:user
sql_passwd:password
sql_refresh_time:600
sql_optimize_clauses:true
sql_history:1d
sql_host:localhost
sql_recovery_logfile:/opt/xx/log/pmacct/recover-daily.log
pidfile:/opt/xx/log/pmacct/daily.pid

sql_history_roundoff:h
plugin_pipe_size: 512000
plugin_buffer_size: 4096
! next must be prime....
sql_cache_entries: 99991

The original pmaccts were set up by someone other than me, so I'm just
trying to figure it out.  Basically the only difference between the two
that I'm seeing is the filter in / out which is there because on #2 the
ethernet we're listening on (eth3) is the internal card and we e only
wanting to record data that is going out of our network, so anything
from the outside world to either out 64.x or the 192.x isn't counted.

I've run tcpdump on both hosts with the filters in it from the pmacct
settings above and the bytes that are caught by tcpdump match properly,
however the values that are written by pmacctd to the respective
databases are consistantly double in #2 :(

Could this be the aggregate[] setting maybe?  I can't see something like
that doubling what pmacctd is recording, but who knows.  Any suggestions
or tips *greatly* appreciated.

Alan


-- 
Alan <[EMAIL PROTECTED]> - http://arcterex.net
--------------------------------------------------------------------
"Backups are for people who don't pray."                 -- big Mike

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

Reply via email to