Richard Daemon wrote:
> Hi all,
>
> Does anyone know how I can go about monitoring bandwidth usage based on
> ports (or service) and maybe client as well?
> I have checked and tried both pfstat and symon and they're both great at
> what they do, but not fully what I'm looking to do.
>
> As for Cacti, I will be trying to get working this weekend in the chroot as
> there's no port yet, unfortunately, but I don't think it will quite do what
> I'm seeking either.
>
> In other words, what I'm looking to do is mainly to monitor and graph the
> average (baseline) bandwidth usage for a few systems and to know what ports
> are mostly used (ftp, http, https, ssh, etc.) and how much bandwidth they
> consume or need, on average.
>
> By doing this, I can also better adjust my ALTQ rules accordingly.
>
> TIA.
>
>   
I would suggest you look at the new snmpd stuff in -current. I don't
follow the tree closely enough to know how far reyk@ and co have got
with it, but I believe that the plan is for it to be well integrated
with pf and the other subsystems in base. Using snmpd to give you a
datasource, you could then use {cacti, mrtg, cricket, $others} to draw
whatever graphs you need.
Si1entDave
(Looking forward to the day spamd and altq info in snmpd is just putting
a few lines in rc.conf.local :-)

Reply via email to