[LARTC] Strange HFSC behavior - bug? my misconfiguration/lack of knowledge?

[Marcin Kałuża]

Hi
I'm moving to hfcs (mostly succesfuly:) ), but I've encountered a strange 
problem: When I download something from this server using wget (one thread) 
via an idle uplink, I get no more than 130 kbit/s, and when I'm using 
prozilla (or start more wgets), I get about 200 kbit, but still it's not all 
the link can do (240) - the rest remamins unused. 
Is this my fault or the qdisc? (I must tell that htb works fine - I get all 
the bandwidth I can no matter how many threads I use)

I enclose my config below

---------------------------cut---------------------------
$uplink=240kbit 
tc qdisc add dev imq0 root handle 1: hfsc default 1009
#main class
tc class add dev imq0 parent 1:0 classid 1:1000 hfsc ls rate $uplink \
	ul rate $uplink
#default
tc class add dev imq0 parent 1:1000 classid 1:1009 hfsc ls rate 1kbit \
	ul rate 10kbit

 
#traffic from the server
tc class add dev imq0 parent 1:1000 classid 1:1001 hfsc ls rate $uplink ul 
rate $uplink
tc filter add dev imq0 parent 1: protocol ip prio 2 u32 \
                match ip src $inetaddr \
                flowid 1:1001

#ssh - low delaay
tc class add dev imq0 parent 1:1001 classid 1:1011 hfsc rt m1 100kbit d 15ms 
m2 30kbit ls rate $uplink ul rate $uplink
tc filter add dev imq0 parent 1:1001 protocol ip prio 2 u32 match ip sport 22 
0xffff flowid 1:1011
tc filter add dev imq0 parent 1:1001 protocol ip prio 2 u32 match ip dport 22 
0xffff flowid 1:1011

#rest from the server
tc class add dev imq0 parent 1:1001 classid 1:1013 hfsc ls rate $uplink ul 
rate $uplink
tc filter add dev imq0 parent 1:1001 protocol ip prio 3 u32 \
                match ip src $inetaddr/32 \
                flowid 1:1013

#lan users 

tc class add dev imq0 parent 1:1000 classid 1:1002 hfsc ls rate 150kbit ul 
rate 150kbit

tc filter add dev imq0 parent 1: protocol ip prio 2 u32 \
                match ip src $lanaddr.0/24 \
                flowid 1:1002