IPFW DUMMYNET: Several pipes after each other
smithi at nimnet.asn.au
Tue Jan 27 23:51:23 PST 2009
On Tue, 27 Jan 2009, Sebastian Mellmann wrote:
> Ian Smith wrote:
> > 00060: 192.000 Kbit/s 0 ms 30 KB 1 queues (1 buckets) droptail
> > 0 tcp 192.168.0.64/1032 220.127.116.11/1863 1847947 563209421 0
> > 0 141
> > 00070: 3.072 Mbit/s 0 ms 40 KB 1 queues (1 buckets) droptail
> > 0 tcp 18.104.22.168/1863 192.168.0.64/1032 2438211 3075075035 0
> > 0 4550
> > It's nearly all streaming rather than more interactive traffic, so
> > pipe latency isn't so much of a concern. Anyway, I rarely actually
> > catch any traffic still in-queue, which you can stare at for tuning.
Just for reference re KES' message re ping times with a full queue: we
only put established TCP traffic through these pipes; ICMP always, and
UDP so far - unless/until it becomes an issue - are free-flowing here.
> > Also, that's aggregate traffic, not per IP as with your masks (which
> > look maybe wider than necessary, 0x0000ffff covers a /16) so you may
> > wind up with lots of separate queues sharing a pipe, which may look
> > very different. How many hosts, how much memory to spare for each?
> Is there any chance to get the dropped packets for _each_ queue (e.g.
> logged to a file for further investigation)?
> Does ipfw provide something here?
I don't know, I've only seen 'ipfw pipe show' results here. If you have
numbered queues specified too, I guess 'ipfw queue show' would be what
to try; if that's any use you could append results to a file/s by cron?
> I'm mainly doing experiments with different kinds of settings (bandwidth
> limitations, variable delay, dropped packets probability etcpp.) and I
> want to see how many packets are actually dropped by ipfw.
Happy experimenting .. soon you'll be the expert we can all consult :)
If you want to get deeper into it, freebsd-net is the appropriate list.
More information about the freebsd-questions