Patrick McHardy wrote:
tc qdisc add dev ppp0 handle 1:0 root htb
tc class add dev ppp0 classid 1:1 htb rate 220kbit
tc filter add dev ppp0 protocol ip u32 match u32 0 0 classid 1:1
gives a 3 pkt queue I think each class gets 3 if there are more classes
That is because you use it on a ppp device and the default txqueuelen
is 3. I guess it would make sense to increase the queue len of the
default queues with HTB (and HFSC), I'm just not sure how what value
would be appropriate. txqueuelen * 10? That would be quite a lot on
ethernet devices. Ideally it should depend on the fraction of the
reserved bandwidth ..
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
3pkt being the default is the point - I guess no one complains because
there is normally another huge buffer behind it. VLANs IIRC have 0 as
default and other tunnel type interfaces maybe - I haven't used them.
If HTB bothers to "fix" these then I think it should just fix them to
something that actually works - maybe 16 or 32 like IMQ/IFB.
I agree it would be nice to work on rates.
It would be nice if htb/hfsc had a variable master rate and the ability
to have a fixed qlen independant of No. of classes :-)
Refering to the above has anyone ever read anything about ingress
shaping on a variable speed link? My dsl is currently 2mbit - 600kbit
due to exchange contention.
With reference to hysteresis 0 - could one of you guys ask Devik
(thinking that it may get read and replied to quicker than if I do it).
Andy.
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html