Skip to contentSkip to conversation

Packets Redirection

LARTC Tom 3 participants 5 messages Oct 24, 2002

lore
Share & tools
lore

Tom

author
#0 lore raw

Hi all,

I use a Debian Woody with Kernel 2.4.18 patched with the HTB queue. I use the tc binay given within the HTB targz. Iptables 1.2.6a
I wrote this script based on the ADSL Bandwith management howto :

DEV=eth0
RATEUP8

tc qdisc add dev $DEV root handle 1: htb default 22
tc class add dev $DEV parent 1: classid 1:1 htb rate ${RATEUP}kbit tc class add dev $DEV parent 1:1 classid 1:20 htb rate $[$RATEUP/7]kbit ceil ${RATEUP}kbit prio 0
tc class add dev $DEV parent 1:1 classid 1:21 htb rate $[$RATEUP/7]kbit ceil 88kbit prio 1
tc class add dev $DEV parent 1:1 classid 1:22 htb rate $[$RATEUP/7]kbit ceil 88kbit prio 2

tc qdisc add dev $DEV parent 1:20 handle 20: sfq perturb 10
tc qdisc add dev $DEV parent 1:21 handle 21: sfq perturb 10
tc qdisc add dev $DEV parent 1:22 handle 22: sfq perturb 10
tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 20 fw flowid 1:20 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 21 fw flowid 1:21 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 22 fw flowid 1:22

iptables -t mangle -N MYSHAPER-OUT
iptables -t mangle -I POSTROUTING -o $DEV -j MYSHAPER-OUT iptables -t mangle -A MYSHAPER-OUT -p tcp --dport 20 -j MARK --set-mark 22 iptables -t mangle -A MYSHAPER-OUT -p icmp -j MARK --set-mark 20 iptables -t mangle -A MYSHAPER-OUT -p udp -j MARK --set-mark 21

I transfered a file by ftp. The packets was sent through the class 22 and its bandwith limited to 88kbits, which is normal. But i sent pings and thoose packets were sent through the class 22 too which is not normal due to the mangle packets marking.
I know which class was used by looking the tc classes status. Why icmp packets was not sent through class 20 ?

Any help would be appriciated. Thanks and have a nice day.

_______________________________________________
LARTC mailing list / [email protected]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

#1 lore raw
Quoted 3 lines · Tom

tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 20 fw flowid 1:20 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 21 fw flowid 1:21 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 22 fw flowid 1:22

This odd. With my setup (2.4.20pre10 and tc from the 3.6 tarball) I need to specify classid instead of flowid:

# tc filter add fw help
Usage: ... fw [ classid CLASSID ] [ police POLICE_SPEC ]
       POLICE_SPEC := ... look at TBF
       CLASSID := X:Y
Quoted 2 lines

iptables -t mangle -N MYSHAPER-OUT
iptables -t mangle -I POSTROUTING -o $DEV -j MYSHAPER-OUT

Tried OUTPUT chain for locally generated packets?

Quoted 2 lines

I transfered a file by ftp. The packets was sent through the class 22 and its bandwith limited to 88kbits, which is normal.

Yes if normal = default.

Quoted 4 lines

But i sent pings and thoose packets were sent through the class 22 too which is not normal due to the mangle packets marking. I know which class was used by looking the tc classes status. Why icmp packets was not sent through class 20 ?

To see which chain you're required to use, append your MYSHAPER-OUT chain to all default chains in the mangle table and see if it works. Then have a look at the packet counts and remove the unused.

Walter

_______________________________________________
LARTC mailing list / [email protected]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

#2 lore raw
In reply to Tom
Quoted 6 lines · Walter Haidinger
Quoted 4 lines · Tom

tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 20 fw flowid 1:20 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 21 fw flowid 1:21 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 22 fw flowid 1:22

Quoted 7 lines

This odd. With my setup (2.4.20pre10 and tc from the 3.6 tarball) I need to specify classid instead of flowid:

# tc filter add fw help
Usage: ... fw [ classid CLASSID ] [ police POLICE_SPEC ]
       POLICE_SPEC := ... look at TBF
       CLASSID := X:Y

from f_fw.c :

                if (matches(*argv, "classid") == 0 ||
                    matches(*argv, "flowid") == 0) {

So both can be used.

Quoted 4 lines
Quoted 2 lines

iptables -t mangle -N MYSHAPER-OUT
iptables -t mangle -I POSTROUTING -o $DEV -j MYSHAPER-OUT

Tried OUTPUT chain for locally generated packets?

You can use OUTPUT and POSTROUTING. See www.docum.org under KPTD.

Stef

Signature · 8 lines
[email protected]
 "Using Linux as bandwidth manager"
     http://www.docum.org/
     #lartc @ irc.oftc.net

_______________________________________________
LARTC mailing list / [email protected]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
author
#3 lore raw
In reply to Tom

Rehi all,

I continue to search and now it's work. There is what i did :
- I change my interface from eth0 to ppp0
- I change the chain from POSTROUTING to OUTPUT (evene if the both possibly work)
- I add the quantum to the SFQ queue, it was 1492 now it's 1500 (the same as my MTU)
- I reduce the bandwith from 128kbps to 104kbps according to the howto
- I change priority of all the handle from 0 to respectivly 0,1,2,3,4,5 and 6 and I changed the flowid to classid.

Now, i'm going to try back my old config step by step to see what exactly made problem.

Thanks a lot for your advises.

Best regards, Tom.

Quoted 30 lines · you
Quoted 6 lines · Walter Haidinger
Quoted 4 lines · Tom

tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 20 fw flowid 1:20 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 21 fw flowid 1:21 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 22 fw flowid 1:22

Quoted 7 lines

This odd. With my setup (2.4.20pre10 and tc from the 3.6 tarball) I need to specify classid instead of flowid:

# tc filter add fw help
Usage: ... fw [ classid CLASSID ] [ police POLICE_SPEC ]
       POLICE_SPEC := ... look at TBF
       CLASSID := X:Y

from f_fw.c :

                if (matches(*argv, "classid") = 0 ||
                    matches(*argv, "flowid") = 0) {

So both can be used.

Quoted 4 lines
Quoted 2 lines

iptables -t mangle -N MYSHAPER-OUT
iptables -t mangle -I POSTROUTING -o $DEV -j MYSHAPER-OUT

Tried OUTPUT chain for locally generated packets?

You can use OUTPUT and POSTROUTING. See www.docum.org under KPTD.

Stef

--

Quoted 8 lines

[email protected]
"Using Linux as bandwidth manager"
http://www.docum.org/
#lartc @ irc.oftc.net

_______________________________________________
LARTC mailing list / [email protected]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

_______________________________________________
LARTC mailing list / [email protected]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

#4 lore raw
In reply to Tom
Quoted 9 lines · Tom

Hi all,

I use a Debian Woody with Kernel 2.4.18 patched with the HTB queue. I use the tc binay given within the HTB targz. Iptables 1.2.6a
I wrote this script based on the ADSL Bandwith management howto :

DEV=eth0
RATEUP=128

Quoted 8 lines

tc qdisc add dev $DEV root handle 1: htb default 22
tc class add dev $DEV parent 1: classid 1:1 htb rate ${RATEUP}kbit tc class add dev $DEV parent 1:1 classid 1:20 htb rate $[$RATEUP/7]kbit ceil ${RATEUP}kbit prio 0
tc class add dev $DEV parent 1:1 classid 1:21 htb rate $[$RATEUP/7]kbit ceil 88kbit prio 1
tc class add dev $DEV parent 1:1 classid 1:22 htb rate $[$RATEUP/7]kbit ceil 88kbit prio 2

a tip : make sure the sum of the rates = ceil of the parent. It's not needed,
but it's more logic. And you have to be sure YOU are the bottleneck on the
link. It's possible that you have to set RATE=120 to get results.

Quoted 14 lines

tc qdisc add dev $DEV parent 1:20 handle 20: sfq perturb 10
tc qdisc add dev $DEV parent 1:21 handle 21: sfq perturb 10
tc qdisc add dev $DEV parent 1:22 handle 22: sfq perturb 10
tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 20 fw flowid 1:20 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 21 fw flowid 1:21 tc filter add dev $DEV parent 1:0 prio 0 protocol ip handle 22 fw flowid 1:22

iptables -t mangle -N MYSHAPER-OUT
iptables -t mangle -I POSTROUTING -o $DEV -j MYSHAPER-OUT iptables -t mangle -A MYSHAPER-OUT -p tcp --dport 20 -j MARK --set-mark 22 iptables -t mangle -A MYSHAPER-OUT -p icmp -j MARK --set-mark 20 iptables -t mangle -A MYSHAPER-OUT -p udp -j MARK --set-mark 21

Quoted 6 lines

I transfered a file by ftp. The packets was sent through the class 22 and its bandwith limited to 88kbits, which is normal. But i sent pings and thoose packets were sent through the class 22 too which is not normal due to the mangle packets marking.
I know which class was used by looking the tc classes status. Why icmp packets was not sent through class 20 ?

Can you check out the iptables counters to see if the packets are really marked like you want ?
And you can get the same filter results if you add a fw filter with no option. The mark will be used as classid :
tc filter add dev $DEV parent 1:0 prio 0 protocol ip fw

Stef

Signature · 8 lines
[email protected]
 "Using Linux as bandwidth manager"
     http://www.docum.org/
     #lartc @ irc.oftc.net

_______________________________________________
LARTC mailing list / [email protected]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/